Models
Every model below is called through the same OpenAI-compatible endpoint: pass
its id as model in a chat completion. Prices come
from the same table billing reads, which is refreshed daily; this page is a
copy of that table, checked against it every week. For the current list at any
moment, call GET /v1/models — each entry carries pricing.input and
pricing.output in the same units.
28 models. Prices are US dollars per million tokens, checked against the API every week and last changed on 9 September 2026.
| Model id | Input $/M | Output $/M |
|---|---|---|
deepseek-ai/DeepSeek-R1-0528 | 0.625 | 2.51 |
deepseek-ai/DeepSeek-V3.2 | 1.63 | 2.92 |
deepseek-ai/DeepSeek-V4-Flash | 0.219 | 0.563 |
deepseek-ai/DeepSeek-V4-Pro | 1.92 | 3.84 |
google/gemma-4-26b-a4b-it | 0.117 | 0.405 |
Intel/Qwen3-Coder-480B-A35B-Instruct-int4-mixed-ar | 0.489 | 2.36 |
meta-llama/Llama-3.3-70B-Instruct | 0.667 | 1.14 |
meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | 0.301 | 0.989 |
MiniMaxAI/MiniMax-M2.5 | 0.323 | 1.29 |
MiniMaxAI/MiniMax-M2.7 | 0.504 | 1.88 |
mistralai/Mistral-Nemo-Instruct-2407 | 0.070 | 0.109 |
moonshotai/Kimi-K2-Instruct-0905 | 0.627 | 2.53 |
moonshotai/Kimi-K2-Thinking | 0.660 | 2.75 |
moonshotai/Kimi-K2.5 | 0.614 | 3.23 |
moonshotai/Kimi-K2.6 | 0.776 | 3.48 |
moonshotai/Kimi-K2.7-Code | 1.18 | 5.12 |
openai/gpt-oss-120b | 0.198 | 0.671 |
openai/gpt-oss-20b | 0.056 | 0.224 |
Qwen/Qwen3-Next-80B-A3B-Instruct | 0.129 | 1.25 |
Qwen/Qwen3.6-27B | 0.377 | 3.31 |
Qwen/Qwen3.6-35B-A3B | 0.206 | 1.37 |
zai-org/GLM-4.5-Air | 0.172 | 1.03 |
zai-org/GLM-4.6 | 0.590 | 2.28 |
zai-org/GLM-4.7 | 0.968 | 2.61 |
zai-org/GLM-4.7-Flash | 0.069 | 0.440 |
zai-org/GLM-5 | 0.946 | 3.06 |
zai-org/GLM-5.1 | 1.52 | 4.79 |
zai-org/GLM-5.2 | 1.74 | 5.47 |