Model list
477 models, and 38 ways to stop choosing one
The public endpoint at /v1/models returns 477 entries: 317 text and
multimodal models, plus 160 image-generation models. Thirty-eight of those are
combination routes — you name the kind of work, and the router picks a specific model
at request time. The remaining 456 map to one model each.
Combination routes
These are the part a plain model list does not have. Instead of tracking which model is
best at code this month, you send auto/best-coding and the choice happens
server-side.
What you give up is knowing which model answered. A route promises an intent, not a model. If you need to reproduce a result exactly, or name a model for compliance reasons, send a specific id instead. Routes are for when you want the right answer more than you want to know where it came from.
| Route | What the name says it selects for |
|---|---|
auto/best-coding | best available model for code |
auto/best-reasoning | best available model for reasoning |
auto/best-vision | best model that accepts images |
auto/best-chat | best general conversation model |
auto/best-fast | fastest acceptable model |
auto/best-coding-fast | code quality, weighted toward speed |
auto/best-free | best model that still has free quota |
auto/pro-coding | code, top tier |
auto/pro-reasoning | reasoning, top tier |
auto/pro-vision | vision, top tier |
auto/pro-chat | chat, top tier |
auto/pro-fast | speed, top tier |
auto/coding:fast | code, speed-weighted |
auto/coding:cheap | code, cost-weighted |
auto/coding:free | code, free quota only |
auto/coding:pro | code, top tier (128k context) |
auto/coding:reliable | code, uptime-weighted |
auto/reasoning:pro | reasoning, top tier (128k context) |
auto/coding | code, default weighting |
auto/reasoning | reasoning, default weighting |
auto/vision | vision, default weighting |
auto/chat | chat, default weighting |
auto/fast | speed, default weighting |
auto/cheap | cost, default weighting |
auto/smart | capability, default weighting |
auto/offline | models reachable without an external provider |
auto/multimodal | any input or output modality |
auto/claude-opus | the Claude Opus line |
auto/claude-sonnet | the Claude Sonnet line |
auto/gemini | the Gemini line (1M context) |
auto/glm | the GLM line (128k context) |
auto/zai | Z.ai models (128k context) |
auto/gemma | the Gemma line (128k context) |
auto/llama | the Llama line (128k context) |
auto/minimax | the MiniMax line |
auto/mimo | Xiaomi MiMo (1M context) |
The exact model each route resolves to is decided at request time, so it can change as providers add, retire, or reprice models.
What the models can do
Capability counts across the 317 text and multimodal models, read from the same endpoint:
| Capability | Models |
|---|---|
| Extended thinking | 137 |
| Adjustable temperature | 38 |
| Vision (image input) | 93 |
| Tool calling | 302 |
302 of the 317 text models support tool calling, and 232 support reasoning. That is a lower bar than it sounds, but it does mean you can point an agent at most of them without checking first.
The image-generation models carry no context window or reasoning flag, which is why the counts above cover the text models only.
How to see the list yourself
The endpoint is public. No key, no account, no rate limit worth mentioning.
# every model, as JSON
curl -s https://freemodel.online/v1/models
# just the combination routes
curl -s https://freemodel.online/v1/models | grep -o '"auto/[^"]*"'
# how many there are, without trusting this page
curl -s https://freemodel.online/v1/models | grep -o '"id"' | wc -lWhich ones are free
Free quota comes and goes as providers run promotions. See which models have some right now →