Magic Sites · Aura Media Studios
Magic Sites is where it starts. Services are where it scales.
Your site is a front door. What happens once traffic arrives — whether you show up in AI answers, whether leads get routed, whether knowledge gets retrieved, whether humans get replaced with agents — is where real leverage lives. We ship that layer, too. Pick what you need.
magic-sites
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
magic-llm
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
magic-llm-gateway
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
magic-agent
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
magic-workshop
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
magic-loops
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
magic-evolve
Magic Evolve
Your agents get smarter every week. Automatically.
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
magic-video
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
magic-sovereign
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Pricing - magic-sites
| Spark | $5/mo | 1 site, 500 chat msgs, community support |
| Studio | $19/mo | 3 sites, 5K chat msgs, custom domain |
| Agency | $99/mo | 25 sites, 50K chat msgs, white-label |
Pricing - magic-llm
| Starter pack | $25 | First-key trial, 12-month expiry |
| Builder pack | $100 | Solo dev + light prod |
| Team pack | $500 | SMB-scale production app |
| Scale pack | $1,000 | Agency / high-volume |
Pricing - magic-llm-gateway
No pricing table for magic-llm-gateway
Pricing - magic-agent
| Starter | $19/mo | 5,000 msgs / KV memory |
| Pro | $49/mo | 50,000 msgs / D1 + Vectorize / MCP |
| Scale | $149/mo | Unlimited msgs / R2 / priority |
Pricing - magic-workshop
| Solo | $39/mo | 1 vCPU / 2GB / 1 repo |
| Pro | $99/mo | 2 vCPU / 4GB / 5 repos / terminal + gateway |
| Team | $299/mo | 4 vCPU / 8GB / unlimited repos / multi-user |
| Enterprise | Custom | Per-seat quote |
Pricing - magic-loops
| Starter | $29/mo | 5 workflows / 1K runs |
| Pro | $79/mo | 50 workflows / 50K runs / marketplace publish |
| Scale | $249/mo | Unlimited workflows / runs / featured |
Pricing - magic-evolve
| Starter | $19/mo | 1M tokens / memory consolidation |
| Pro | $79/mo | 10M tokens / preference + skill learning |
| Scale | $249/mo | 100M tokens / autopsy + A/B promotion |
Pricing - magic-video
| Free | $0 | 5 videos / 10GB delivery |
| Starter | $9/mo | Branded player / 100GB |
| Pro | $29/mo | Heatmaps + A/B / 500GB |
| Scale | $99/mo | White-label / API / multi-tenant / unlimited delivery |
Pricing - magic-sovereign
No pricing table for magic-sovereign
Sales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSales touch
Get a quote
Tell us agent count, est tokens / mo, deployment preference, use case. Mike + Brooke follow up within two business days.
Request quoteSide-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Side-by-side
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Sites
Websites with AI built in. Cheaper than WordPress hosting.
$5/mo
- - Edge-hosted on Cloudflare, scale-to-zero
- - Chat-driven editor + agent-native
- - 500 chat messages bundled on Spark
- - Pay-per-use beyond the bundle, same runtime on every tier
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic LLM
One API key for every top open-weights model. OpenAI-compatible.
$25 credit pack
- - Curated shelf - top model per use case
- - Pay-per-token from a $25 starter credit pack
- - Llama 3.3 70B, Gemma 4, Kimi K2.6, Qwen Coder, Whisper, Flux
- - OpenAI-compatible endpoint, no rewrite to switch
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic LLM Gateway
BYOK routing for your own Anthropic / OpenAI / Gemini keys.
Pricing TBA
- - Failover, fallback, and cost routing across providers
- - Observability - per-call cost, latency, hit rate
- - Guardrails - PII redaction, prompt-injection filters
- - One endpoint instead of five SDKs
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Agent
Drop-in chat / AI for your site. Installed in minutes.
$19/mo
- - Scale-to-zero edge runtime (Worker + Durable Object)
- - Per-site memory - KV on Starter, D1 + Vectorize on Pro
- - MCP tool support on Pro / Scale
- - Embeddable widget or full takeover
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Workshop
Always-on AI dev environment. Browser IDE + terminal.
$39/mo
- - 1-4 vCPU containers with FUSE-mounted R2
- - Multi-repo, persistent
- - Solo $39, Pro $99, Team $299
- - Enterprise quoted per-seat
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Loops
Code-as-tool automation. Replaces n8n / Zapier / Make.
$29/mo
- - Triggers - webhook, cron, email, file, MCP, Discord
- - 1K - unlimited workflows by tier
- - Marketplace publish on Pro / Scale
- - Embedded LLM included on every tier
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Evolve
Your agents get smarter every week. Automatically.
$19/mo
- - STaR / Voyager / Reflexion-style auto-improvement
- - Memory consolidation, preference learning, skill extraction
- - Workflow autopsy + A/B promotion on Scale
- - 1M - 100M tokens / mo by tier
Side-by-side
Magic Sovereign
Private AI on hardware you own. Custom-quoted.
From $5K (single box)
- - Single DGX Spark from $5K (1-3 users, RAG + workflows)
- - Workstation cluster $40K-200K (5-25 users)
- - GPU server $300K-450K (25-200 users)
- - Hardware++ pricing - hardware + setup + license + management
Magic Video
Video hosting with branded player + Magic Interactive AI layer.
Free - $99/mo
- - Free tier - 5 videos, 10GB delivery, basic player
- - Pro $29 - branded, CTAs, A/B test, heatmaps
- - Scale $99 - white-label, API, multi-tenant
- - Magic Interactive - Pause-and-Ask, branching, hot-lead alerts
Free
AI readiness audit
We'll run our stack against your real pages and tell you what's worth turning on - SEO, AEO, agent, RAG, automation. No pitch.
Take the auditMagic LLM model shelf
| Use case | Model | Retail |
|---|---|---|
| General chat | Llama 3.3 70B | $1.50 |
| Fast / cheap | Gemma 4 26B | $0.50 |
| Reasoning | Kimi K2.6 | $2.50 / $8.00 |
| Code | Qwen 2.5 Coder | $1.25 |
| Embeddings | BGE / GTE family | $0.05 |
| Image generation | Flux / SDXL | $0.10 |
| Speech-to-text | Whisper Large v3 | $0.01 |
USD per million tokens unless noted
Magic Sovereign reference deployments
Single box
$4.699K-$5KSolo / small team - 1-3 users
1× Nvidia DGX Spark (128GB unified, 1 PFLOP FP4)
Best for: private RAG, embeddings, document workflows, knowledge agents, local fine-tunes
Workstation cluster
$40K-$200KSmall team / department - 5-25 users
4-8× pro GPUs OR 1× DGX Station-class
Best for: team RAG + workflow automation + light coding swarm + customer support tier-1
GPU server
$300K-$450KDepartment / mid-market - 25-200 users
1× 8-GPU H100 / B100 SXM rack-mount server
Best for: engineering swarm, support replacement, content + research at scale, data agents
Single rack
$2.0M-$5.0MEnterprise - 200-2000 users
Multi-server rack (4-8× nodes + InfiniBand + cooling)
Best for: full enterprise agent platform, cross-BU RAG, quant signal extraction
Multi-rack
$10,000K+Hyperscale / custom - 2000+ users
Custom architecture, multiple racks, bespoke networking
Best for: AI-first companies running every function on agents
Hardware++ = hardware + setup + license + management. Quote-based.
Five ways to partner
- ResellerSell our stack white-labeled. Paid in ARC credits.
- AffiliateRefer customers, earn revenue share on every renewal.
- IC SalesIndependent sales contractor. Carry the bag, keep the commission.
- Agency partnerBuild on Magic Sites for clients. Margin on services.
- Tech partnerPlug your tool into the stack. Co-marketing + listing.
The differentiator
The 12-month rule
We don't sell five-year contracts. The AI landscape ships a new winner every 4-12 weeks.
- - Re-curate the model shelf quarterly
- - Swap providers when the math changes
- - Refresh architecture before it goes stale
- - Customers walk away with everything they built. No lock-in.
The flywheel
Magic Sites -> traffic Magic Agent -> conversations Magic LLM -> tokens Magic Loops -> automations Magic Evolve -> better agents back to Magic Sites
Every product feeds the next. The site brings traffic. The agent captures conversations. The LLM runs the agent. Loops automates work the agent identifies. Evolve makes every agent smarter from what it learned. The smarter agent draws better traffic. Repeat.
Watch this part
Watch this part
Ask this video
Try asking
// INTERACTIVE_VIDEO_AGENT · V1_PROTOTYPE
Talk to the video. It shows you what you want.
Type a question or hit the mic and ask out loud. The agent picks the right card from a deterministic content toolbox and shows it next to the video. Built on the same pattern that powers Magic Stream + Magic Agentic Video. Try: "what is Magic Sites?", "show me pricing", "I need an audit", "join your discord".
Launch a real site
Start free. Edit it. Put it online.
Create an account and Magic Sites launches a basic editable site on a Magic Sites address. No card and no checkout. Upgrade when you need more pages, automation, higher AI limits, or agency tools.
- ✓ Real hosted starter site
- ✓ Secure WorkOS account and editor access
- ✓ Magic Sites subdomain
- ✓ Domain registration, renewal, transfer, and connection remain separate
Pricing
Cheaper than WordPress hosting. AI built in.
Magic Sites starts at $5/mo per site — less than most shared WordPress plans, and your site already comes with a working AI chat agent. Scale by usage, not by upgrading. The $5/mo Spark rate IS the limited-time offer (no further founding discount on Spark); founding rates lock for the life of your account: first 50 Studio at $15/mo, first 10 Agency at $79/mo.
Spark
Your first AI-native website. Cheaper than shared WordPress hosting, and the AI is already wired up.
vs Bluehost / SiteGround entry plans
- ✓ 1 Magic Site
- ✓ Our EmDash-based template engine
- ✓ 500 chat messages / month included
- ✓ Pay-per-use after that - no upgrade required
- ✓ 1GB media storage
- ✓ Full plugin runtime (community support)
Studio
Most popularThree sites + connect a customer-owned domain + bigger AI quota. For small teams shipping multiple brands or clients.
vs DreamPress / WP Engine entry
- ✓ 3 Magic Sites
- ✓ Connect 1 customer-owned domain (domain registration and renewal separate)
- ✓ 5,000 chat messages / month included
- ✓ Plugin marketplace + custom installs
- ✓ 10GB media storage
- ✓ Email priority support + WorkOS SSO
Agency
Twenty-five sites + white-label + reseller. For agencies billing their own clients.
vs Kinsta agency / WP Engine reseller
- ✓ 25 Magic Sites
- ✓ Connect customer-owned domains on up to 25 sites (domain registration and renewal separate)
- ✓ White-label admin + your branding
- ✓ Stripe Connect - resell at your markup
- ✓ 50,000 pooled chat messages / month
- ✓ 100GB media storage
- ✓ Dedicated support channel
// PAY_PER_USE
Same prices on every tier. Pay only for what you use beyond your bundle.
We don't gate features by tier. Need more chat? More agents? More image generation? Same per-unit cost whether you're on Spark or Agency. Tiers only differ on the bundle that's included up front.
Fast open-source model. Spark includes 500/mo, Studio 5K, Agency 50K.
Frontier-grade reasoning when you need it.
Hero images, social posts, product mocks — on demand.
First 1K runs/mo free on every tier.
Lead-qual / FAQ / SDR / support / closer — pick what you need.
Connection entitlement only. Domain registration, purchase, renewal, and transfer fees are never included.
Services shelf
Beyond the site — services that compound.
SEO Automation
from $1,500/mo
Your SEO on autopilot — we surface, you decide.
Our seo-intel worker crawls your site, your competitors, and the SERPs weekly. You get monthly strategy — not a 40-page PDF nobody reads.
Start SEO →AEO Monitoring
from $495/mo
See yourself the way ChatGPT sees you.
ChatGPT. Claude. Perplexity. Gemini. We watch what they say about you, and tell you how to change it.
Start AEO →AI Agents + Automation
from $18,500 project + $2,500/mo ops
Every manual step in your business — automated, agented, accountable.
The manual process costing you $10k/mo in human time? Agented in weeks, not quarters. Edge-hosted, secure auth, yours forever.
Scope an automation →Custom RAG
from $22,500 project + $950/mo ops
Turn your documents into an agent your team can talk to.
Your docs, your data, your private agent. GraphRAG by default. Enterprise CNA-powered option available under NDA.
Scope a knowledge agent →SiteOps Consulting
from $1,500/mo
An operator-in-residence for your AI-native business.
Operator-in-residence for founders who want thinking partners, not vendors. Architecture reviews, deal reviews, vendor reviews, monthly calls.
Book SiteOps →Every Magic Sites tenant gets a 10% retainer discount on their first services engagement. Bundles below.
Bundle offers
Stack the layer. Discount the math.
Entry Bundle
Magic Sites + SEO
Magic Sites tier + SEO Entry retainer
Magic Sites tier rate + $1,200/mo
SEO normally $1,500/mo — 20% bundle discount
“Ship your site and we'll rank it.”
Talk bundles →AI-Era Bundle
Magic Sites + AEO
Magic Sites tier + AEO Entry
Magic Sites tier + $395/mo
AEO normally $495/mo — 20% bundle discount
“Own your AI answer surface from day one.”
Talk bundles →All-Five Bundle
Full AMS Stack
Magic Sites (any tier) + SEO + AEO + Agents project + SiteOps consulting + any Custom RAG project
$4,950/mo
Sum at Entry tiers $6,945/mo — 29% discount · 12-month minimum
“Your outsourced growth + ops stack. One vendor, one invoice, one throat to choke.”
Talk bundles →// FREE
We track AI weekly. You get the highlights.
New model launches, what's actually worth using, what to ignore, and the rare changes that affect your stack. One short email, no fluff, unsubscribe with one click.
→ What changed in AI this week (the parts that matter)
→ One playbook you can use this week
→ A model that's worth your attention right now
→ Free AI readiness score on signup
// FREE_AUDIT
Free AI Readiness Audit
Five questions, instant score, plus a one-page report on where to start. No sales call required.
Take the audit →