kitio.aiAdded Oct 28, 2026
Avian
PaidOpenAI-compatible inference API for open models, plus dedicated B200 nodes
What is Avian?
Avian is a pay-per-token inference API for open models such as DeepSeek, MiniMax, GLM, Kimi and MiMo, served through an OpenAI-compatible endpoint. Developers who need more capacity can also reserve dedicated NVIDIA B200 and B300 GPU nodes.
Switching from OpenAI takes a change of base URL and key, and the API works with coding tools like Claude Code, Cursor, Cline and Kilo Code. Models run on NVIDIA B200 GPUs with speculative decoding, and the service states it has no rate limits and stores no data.
Avian Compute rents whole 8-GPU HGX nodes on prepaid terms from 12 months to three years, priced per GPU hour. Token usage is paid from prepaid credits, and infrastructure runs on Microsoft Azure.
Features
- OpenAI-compatible API
- Open model catalog
- Prepaid token billing
- Dedicated B200 and B300 nodes
- Coding tool support
Tags
Releases
+ Submit new release- Listed
Avian was added to the directory.
Organization
Avian
Listed by kitio.ai · unverified owner
Pricing
- Pricing model
- Paid
- Paid options from
- Not listed
Avian plans
Serverless API
From $0.1275 / 1M input tokens
- Pay per token
- Prepaid credits
- All hosted models
Dedicated B200 & B300
From $3.42 / GPU-hour
- Full 8-GPU HGX nodes
- Prepaid 12-month to 3-year terms
Prices in USD, checked Oct 29, 2026. Plans change often, so confirm on the Avian site.Official pricing
Sponsored
Your AI tool here
A Featured launch puts your tool in this spot and on the home page for a week.
Top Avian alternatives
View allModels & APIs
Serverless inference for open models on Cloudflare's global network
Models & APIs
IBM's enterprise studio for Granite and third-party models
Models & APIs
Production inference and fine-tuning for open models on Nebius cloud
Models & APIs
Fast inference for large open models on SambaNova's RDU chips
Compare Avian
If you liked Avian
Inference cloud for open models like DeepSeek, Qwen and GLM
FreemiumModels & APIsMade in ChinaCompute, RL training, evals and inference for building your own models
PaidModels & APIsMade in United StatesSmall task-specific open models for search, extraction and classification
PaidModels & APIs- TextSynthNew
REST API and playground for open language and image models
FreemiumModels & APIsMade in France
Pros and cons
Pros
- Low per-token prices
- Drop-in OpenAI replacement
- Option to reserve whole GPU nodes
Cons
- Mostly open models, no GPT or Claude
- Credits must be prepaid
- Compute requires long-term terms
Prompts & results
No prompts yet. Add your own prompt and output to help others use this tool.
Share a prompt
Reviews
Q&A
What is Avian?
Avian is a pay-per-token inference API for open models such as DeepSeek, MiniMax, GLM, Kimi and MiMo, served through an OpenAI-compatible endpoint. Developers who need more capacity can also reserve dedicated NVIDIA B200 and B300 GPU nodes.
Can I use Avian with Claude Code or Cursor?
Yes. Avian lists support for more than 20 coding tools, including Claude Code, Cursor, Cline and Kilo Code.
How much does Avian cost?
Avian is paid.
Is there a free version of Avian?
No. Avian is a paid tool.
What are the main features of Avian?
OpenAI-compatible API, Open model catalog, Prepaid token billing, Dedicated B200 and B300 nodes, Coding tool support.
Who makes Avian?
Avian is made by Avian.
What are the best alternatives to Avian?
Popular alternatives include Cloudflare Workers AI, IBM watsonx.ai, Nebius Token Factory.