Cheapest reasoning model API
The cheapest reasoning model API on OceanToken is Qwen3.7 Flash, at $0.03 / $0.13 per 1M tokens. Below, 20 models ranked by the price every account pays.
Ranked by price 20
| # | Model | Family | Context | Price | Discount |
|---|---|---|---|---|---|
| 1 | Qwen3.7 Flashqwen/qwen3.7-flash | Qwen | 984K | $0.03 / $0.13 per 1M tokens | — |
| 2 | Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning | Nemotron | 262K | $0.06 / $0.16 per 1M tokens | — |
| 3 | gpt-oss-20bopenai/gpt-oss-20b | GPT | 131K | $0.04 / $0.18 per 1M tokens | — |
| 4 | Qwen3.5-9Bqwen/qwen3.5-9b | Qwen | 262K | $0.10 / $0.15 per 1M tokens | — |
| 5 | Nemotron 3 Nano 30B A3Bnvidia/nemotron-3-nano-30b-a3b | Nemotron | 262K | $0.06 / $0.24 per 1M tokens | — |
| 6 | Qwen3.5-Flashqwen/qwen3.5-flash-02-23 | Qwen | 984K | $0.065 / $0.26 per 1M tokens | — |
| 7 | Qwen3 14Bqwen/qwen3-14b | Qwen | 41K | $0.12 / $0.24 per 1M tokens | — |
| 8 | Seed 1.6 Flashbytedance-seed/seed-1.6-flash | Seed | 262K | $0.075 / $0.30 per 1M tokens | — |
| 9 | gpt-oss-safeguard-20bopenai/gpt-oss-safeguard-20b | GPT | 131K | $0.075 / $0.30 per 1M tokens | — |
| 10 | Nemotron 3.5 Content Safetynvidia/nemotron-3.5-content-safety | Nemotron | 131K | $0.20 / $0.20 per 1M tokens | — |
| 11 | GPT-5 Nanoopenai/gpt-5-nano | GPT | 400K | $0.045 / $0.36 per 1M tokens | −10% |
| 12 | Gemini 2.5 Flash Litegoogle/gemini-2.5-flash-lite | Gemini | 1.05M | $0.09 / $0.36 per 1M tokens | −10% |
| 13 | GLM 4.7 Flashz-ai/glm-4.7-flash | GLM | 200K | $0.0605 / $0.40 per 1M tokens | — |
| 14 | Nemotron 3 Supernvidia/nemotron-3-super-120b-a12b | Nemotron | 1.05M | $0.085 / $0.40 per 1M tokens | — |
| 15 | Seed-2.0-Minibytedance-seed/seed-2.0-mini | Seed | 262K | $0.10 / $0.40 per 1M tokens | — |
| 16 | GPT-6 Lunaopenai/gpt-6-luna | GPT | 1.05M | $0.09 / $0.45 per 1M tokens | −10% |
| 17 | Qwen3 32Bqwen/qwen3-32b | Qwen | 131K | $0.14 / $0.40 per 1M tokens | — |
| 18 | Gemma 4 31Bgoogle/gemma-4-31b-it | Gemma | 262K | $0.15 / $0.40 per 1M tokens | — |
| 19 | Qwen3.8 Flashqwen/qwen3.8-flash | Qwen | 984K | $0.143 / $0.447 per 1M tokens | −25% |
| 20 | GPT-6 Luna Proopenai/gpt-6-luna-pro | GPT | 1.05M | $0.10 / $0.50 per 1M tokens | — |
About these prices
Frequently asked questions
What is the cheapest reasoning model API?
On OceanToken it is Qwen3.7 Flash, at $0.03 / $0.13 per 1M tokens. Next is Nemotron 3.5 Lightning, at $0.06 / $0.16 per 1M tokens. Both answer on the same API with the same key.
How are the models ranked?
Ranked by blended price (half input, half output tokens, per 1M), as every account pays it; only models of named families are ranked.
Is the cheapest model good enough?
The price and the listed capabilities are facts this page can show; how well a model does your task is not. Run the same prompts on two models: with one key, switching is a one-line change.
Keep exploring
Nemotron API · GPT API · Seed API · Gemini API · Gemma API
Cheapest LLM API · Cheapest long-context LLM API · Cheapest vision LLM API · Cheapest video generation API · Cheapest image generation API · Cheapest embedding API