Esta sección aún no está traducida. El contenido siguiente se muestra en inglés, tal como se publicó. Leer la versión en inglés.
DeepSeek · Text model · GA
Compare with similar models Open the Model Finder
Figures checked 1 Nov 2026. A dash means not published on the sources we checked, so it stays blank rather than estimated. AA index and speed are from Artificial Analysis at the max setting.
DeepSeek names these settings reasoning_effort. Price per token is the same at every level. Higher levels usually use more output tokens. api-docs.deepseek.com (opens in a new tab)
| Level | When to use it, per DeepSeek | AA index | Tokens/s |
|---|---|---|---|
| none | No per-level advice published | 18 artificialanalysis.ai (opens in a new tab) | – |
| low | No per-level advice published | – | – |
| high default | No per-level advice published | – | – |
| max | No per-level advice published | 32 artificialanalysis.ai (opens in a new tab) | 175 artificialanalysis.ai (opens in a new tab) |
Our rule of thumb: Simple or speed-critical work: low. Standard work: the lab's default. Complex work: high. Use xhigh or max only when your own tests show a gain, as OpenAI and Anthropic both advise. developers.openai.com (opens in a new tab)platform.claude.com (opens in a new tab)
Artificial Analysis Intelligence Index at each setting it tested. Levels it has not tested are left out.
| Level | AA index | Tokens/s |
|---|---|---|
| none | 18 | – |
| max | 32 | 175 |
Highlighted point is DeepSeek V4.1 Flash 1017. Grey points are the other 26 text models with both figures. Index at each model's best tested setting.
| Model | Index | Blended US$/1M |
|---|---|---|
| GPT-6 Astra | 57 | 20.000 |
| GPT-6.1 Sol | 55 | 4.000 |
| GPT-6.1 Terra | 51 | 2.000 |
| GPT-6 Luna | 28 | 0.200 |
| Claude Fable 5.1 | 50 | 20.000 |
| Claude Opus 5.5 | 58 | 8.000 |
| Claude Sonnet 5.5 | 54 | 4.000 |
| Claude Haiku 5.5 | 36 | 0.200 |
| Gemini 4 Carbon | 59 | 4.000 |
| Gemini 4 Argon | 50 | 4.000 |
| Gemini 3.8 Flash | 31 | 1.500 |
| Gemini 3.5 Flash-Lite | 12 | 0.850 |
| Grok 4.7 | 35 | 3.000 |
| Muse Wonder | 43 | 8.000 |
| Mistral Large 4 | 26 | 1.050 |
| Mistral Small 4 | 4 | 0.262 |
| DeepSeek V4.1 Flash 1017 | 32 | 0.525 |
| DeepSeek V4 Pro | 23 | 1.980 |
| Qwen3.8-Max | 34 | 3.000 |
| Kimi K3.1 Pro | 46 | 4.000 |
| GLM-5.3 | 32 | 2.150 |
| GLM-5.3-Flash | 28 | 0.237 |
| MiniMax M3.1 Flash | 29 | 0.262 |
| MiniMax M3 | 19 | 0.525 |
| MiMo-V2.6-Pro | 30 | 0.544 |
| MiMo-V2.6-Flash | 27 | 0.175 |
| Step 5 Preview | 29 | 1.425 |
US$ per 1M tokens, standard tier, short prompts. Output left blank where not published.
| Model | Input US$/1M | Output US$/1M |
|---|---|---|
| DeepSeek V4.1 Flash 1017 | 0.3 | 1.2 |
| MiMo-V2.6-Pro | 0.435 | 0.87 |
| MiniMax M3.1 Flash | 0.15 | 0.6 |
| Gemini 3.8 Flash | 0.75 | 3.75 |
| GLM-5.3-Flash | 0.15 | 0.5 |
| Lab | DeepSeek |
|---|---|
| Type | Text |
| Status | GA api-docs.deepseek.com (opens in a new tab) |
| Released | 17 Oct 2026 huggingface.co (opens in a new tab) |
| API access | Yes api-docs.deepseek.com (opens in a new tab) |
| API model ID | – |
| API rate limits | – |
| Consumer app plans | – |
| Price at each level | Same per-token price at every level. api-docs.deepseek.com (opens in a new tab) |
| Input per 1M tokens | US$0.3 api-docs.deepseek.com (opens in a new tab)Peak rate (01:00 to 04:00 and 06:00 to 10:00 UTC weekdays). Half off-peak |
| Output per 1M tokens | US$1.2 api-docs.deepseek.com (opens in a new tab)Peak rate. Half off-peak |
| Cached input | US$0.006 api-docs.deepseek.com (opens in a new tab)Peak |
| Batch in / out | – |
| Other pricing | – |
| Context window | 1M tokens api-docs.deepseek.com (opens in a new tab) |
| Max output | 384K tokens api-docs.deepseek.com (opens in a new tab) |
| Input types | text, image api-docs.deepseek.com (opens in a new tab) |
| Output types | text |
| Weights | Open huggingface.co (opens in a new tab)MIT |
| Reasoning | Yes api-docs.deepseek.com (opens in a new tab)Thinking and non-thinking |
| Tool use | Yes api-docs.deepseek.com (opens in a new tab) |
| Artificial Analysis index | 32 (max) artificialanalysis.ai (opens in a new tab) |
| Output speed | 175 tokens/s artificialanalysis.ai (opens in a new tab) |
| Time to first token | 1.1 s artificialanalysis.ai (opens in a new tab) |
| Reasoning levels | none, low, high (default), max api-docs.deepseek.com (opens in a new tab) |
| Lab benchmark claims |
|
| Limitations |
|
DeepSeek does not document calling MCP servers from its own API. You can still connect tools through an agent or framework. Browse AI tools
Same type, closest on quality index and blended price. A similarity, not a ranking.
Compare DeepSeek V4.1 Flash 1017 with MiMo-V2.6-Pro and MiniMax M3.1 Flash
All 109 AI models · AI agents · How we check sources
DeepSeek 
DeepSeek says V4.1-Flash-1017 finishes tasks faster with shorter answers at unchanged API prices, from $0.15 in and $0.60 out per 1M tokens off-peak. The weights are on Hugging Face under the MIT licence. Benchmark gains are DeepSeek's own figures.
DeepSeek
DeepSeek's Search and Fetch APIs give agents web results and pages as text. Calls made through its own harness, Claude Code or Codex were free until billing began on 1 November 2026, Beijing time. More agent traffic means more inference load on DeepSeek's capacity.
