Official site · 🇺🇸 United States
12 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| gpt-oss-20b | fp4 | $0.03 | $0.13 | 131,072 | 99.8% |
| Granite 4.1 8B | bf16 | $0.05 | $0.10 | 131,072 | 99.9% |
| gpt-oss-120b | fp4 | $0.04 | $0.14 | 131,072 | 99.9% |
| DeepSeek V4 Flash | fp8 | $0.14 | $0.28 | 1,048,576 | 99.3% |
| Gemma 4 31B | bf16 | $0.12 | $0.35 | 262,144 | 99.9% |
| Qwen3.6 35B A3B | fp8 | $0.25 | $1.25 | 262,144 | 100.0% |
| Qwen3.5-35B-A3B | fp8 | $0.25 | $1.25 | 262,144 | 100.0% |
| Qwen3.6 27B | fp8 | $0.60 | $3.60 | 262,144 | 100.0% |
| Kimi K2.7 Code | int4 | $0.94 | $4.00 | 262,144 | 100.0% |
| Kimi K2.6 | fp4 | $0.95 | $4.00 | 262,144 | 97.2% |
| GLM-5.2 | fp4 | $1.39 | $4.40 | 262,144 | 99.9% |
| DeepSeek V4 Pro | fp8 | $1.74 | $3.48 | 1,048,576 | 99.6% |