Router DeskHomeAI model comparison
AI model comparison
Same base URL, different slugs. This table is the routed price from the models API, not a quality ranking.
Updated 2026-09-27
OpenRouter’s compare pages put two model cards side by side. The useful first cut is still price, context, and whether an endpoint is actually up. Quality depends on the task. A coding agent and a short classifier should not share a winner.
| Model | Input / M | Output / M | Context |
|---|---|---|---|
| DeepSeek V4 Flash 0731 | $0.021 | $0.32 | 1.3M |
| DeepSeek V4 Pro 0423 | $0.348 | $0.696 | 1.0M |
| Qwen3 32B | $0.08 | $0.28 | 131K |
| GLM 5.2 | $0.650 | $2.04 | 1.0M |
| Qwen3.7 Max | $1.48 | $4.43 | 1.0M |
| Grok 4.5 | $2 | $6 | 500K |
| Claude Sonnet 5 | $2 | $10 | 1.0M |
| GPT-5.1 | $1.25 | $10 | 400K |
| Claude Haiku 4.5 | $1 | $5 | 200K |
How to read the chart
- Input-heavy retrieval and repo dumps: Flash 0731 or Qwen3 32B.
- Long written answers: V4 Pro 0423 stays under $1 per million output. Sonnet 5 and GPT-5.1 are $10.
- Prompts over 200K tokens: Grok 4.5’s listed rate doubles above that threshold. Sonnet 4.5 does something similar. Sonnet 5’s base row did not show that override.
- A $0 experiment: openrouter/free, which will not be any of the rows above.
OpenRouter can also compare two specific ids, for example Haiku 4.5 against Sonnet 4.5. Use that page when you want the live provider table. Use this one when you are choosing a family.
Questions
Which model is cheapest on OpenRouter?
Among the rows above, DeepSeek V4 Flash 0731 had the lowest input price on 27 Sep 2026. Free slugs are $0 and are a different set.
Does a lower price mean a weaker model?
Not by itself. Flash is a smaller active model than V4 Pro. Haiku is priced below Sonnet and is the one Anthropic describes as the fast tier. Read the card, then run your own prompts.
Where is the official comparison tool?
https://openrouter.ai/compare
Related
- DeepSeek API price
On 27 Sep 2026 OpenRouter listed DeepSeek V4 Pro 0423 at $0.348 / $0.696 per million tokens, and V4 Pro 0813 at $0.245 / $3.50.
- Claude Sonnet 5 pricing
Claude Sonnet 5 was $2 input and $10 output per million tokens on 27 Sep 2026. Batch is $1 and $5. The slug is anthropic/claude-sonnet-5.
- GPT-5.1 API
GPT-5.1 on OpenRouter was $1.25 input and $10 output per million tokens on 27 Sep 2026. The slug is openai/gpt-5.1, with a 400K context.
- GLM 5.2 API
GLM 5.2 on OpenRouter was $0.650 input and $2.042 output per million tokens on 27 Sep 2026. The slug is z-ai/glm-5.2, with a 1M context.