Open. Portable. Observable. Fast.
Every model you and your agent need.One OpenAI-compatible key
Keep your app loop, prompts, tools, structured outputs, evals, and approvals. Swap models, not your agent.
$10/month in free credits for your first 3 months
curl https://tokens.flex.ai/v1/chat/completions \
-H "Authorization: Bearer $FLEXAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "Meta-Llama-3.1-8B-Instruct-FP8",
"messages": [{"role": "user", "content": "Hello from FlexAI"}]
}'The console, from builder to org
The same login, from your first key to org-wide governance.
Same login: builder and org-admin views.
Mistral NemoChatGemma 4 26B A4BChatBGE-M3EmbeddingsNemotron 3 Super 120B A12BChatDeepSeek V4 FlashChatLlama 3.1 8B InstructChatGemma 4 31B ITChatQwen3 Coder 30B A3BChatQwen3.5 9BChatGPT-OSS 120BChatKokoro-82MAudioGLM 5.2ChatWhisper Large V3 TurboTranscriptionLlama 3.3 70B InstructChatQwen3 30B A3B Thinking 2507ChatMiniMax M2.7ChatQwen3 8BChatNVIDIA Parakeet TDT 0.6B v3TranscriptionFLUX.1 [schnell]ImageGLM 4.5 AirChatPaddleOCR-VL 1.5Vision
PaddleOCR-VL 1.5VisionGLM 4.5 AirChatFLUX.1 [schnell]ImageNVIDIA Parakeet TDT 0.6B v3TranscriptionQwen3 8BChatMiniMax M2.7ChatQwen3 30B A3B Thinking 2507ChatLlama 3.3 70B InstructChatWhisper Large V3 TurboTranscriptionGLM 5.2ChatKokoro-82MAudioGPT-OSS 120BChatQwen3.5 9BChatQwen3 Coder 30B A3BChatGemma 4 31B ITChatLlama 3.1 8B InstructChatDeepSeek V4 FlashChatNemotron 3 Super 120B A12BChatBGE-M3EmbeddingsGemma 4 26B A4BChatMistral NemoChat
Results in production
What teams get once they're building on Token Factory.
3 lines
To integrate
Point the OpenAI SDK at FlexAI and swap the model id.
From first call to production
One key the whole way.
- 1Try itAn open, rate-limited demo in the browser. Feel the API before you commit.
- 2IntegrateSwap base_url. Your prompts, tools, RAG, SDKs, and evals carry over unchanged.
- 3Go to productionSpend, limits, and key health are first-class, with reliable error semantics.
- 4
Observe
Routing across models and request metadata, coming to every call.
Frequently Asked Questions
Past steady-state volume? Find your serverless to dedicated break-even
Start with inference. Keep your agent loop
Every open model behind one OpenAI-compatible key, priced at the market rate.
$10/month in free credits for your first 3 months