DeepseekReasoningReleased 21 August 2025
DeepSeek V3.1 pricing, context and benchmark
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates.
164K
Input context
147K
Output context
$0.25
Official input
$0.95
Official output
n/a
AA index
01 · Model Overview
Everything that matters, in one table
Context, pricing, cache, modalities and tooling, from the model registry and Artificial Analysis.
| Model | DeepSeek V3.1 |
| Model ID | deepseek/deepseek-chat-v3.1 |
| Provider | Deepseek |
| Release date | 21 August 2025 |
| Status | Generally available |
| Input context | 163,840 tokens |
| Output context | 147,456 tokens |
| Official input pricing | $0.25 per 1M tokens |
| Official output pricing | $0.95 per 1M tokens |
| Cache read pricing | $0.382 per 1M tokens (28% off input) |
| Market input pricing | $0.25 per 1M (cheapest route) |
| Market output pricing | $0.95 per 1M (cheapest route) |
| Highest route pricing | $0.5 in, $1.5 out per 1M |
| Input modalities | text |
| Output modalities | text |
| Reasoning control | Not exposed |
| Tool support | Text generation only |
| Tokenizer | DeepSeek |
| Provider routes | 4 endpoints |
| Artificial Analysis index | Not published |
| Speed | 74 tokens/s, 2.5s to first token |
| Knowledge cutoff | Not publicly stated |
Input and output support
| Modality | Input | Output | Billed at |
|---|---|---|---|
| Text | ✓ yes | ✓ yes | $0.25 in |
| Image | no | no | not supported |
| Audio | no | no | not supported |
| Video | no | no | not supported |
| File (PDF) | no | no | not supported |
02 · Overview
What is DeepSeek V3.1?
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates.
It extends the DeepSeek-V3 base with a two-phase long-context...
DeepSeek V3.1 carries a 164K token window, uses the DeepSeek tokenizer and is served by Deepseek. On Hoodgen it sits in the Reasoning tier of the model menu next to its siblings, so you can switch mid conversation without losing the thread.
03 · Release
Release and what changed from the previous version
Released21 August 2025
| Property | R1 0528 | DeepSeek V3.1 | Change |
|---|---|---|---|
| Released | 2025 | 2025 | Same release window |
| Context | 164K | 164K | Unchanged |
| Max output | 147K | 147K | Unchanged |
| Input price | $0.5 | $0.25 | Cheaper |
| Output price | $2.15 | $0.95 | Cheaper |
| Input modalities | text | text | Unchanged |
| AA index | n/a | n/a | Newer benchmark run |
04 · Capabilities
Features and capabilities
⚡
Prompt caching
Cached input costs $0.382 per million instead of $0.25, so stable prefixes get cheaper.
07 · Use cases
What can you do with DeepSeek V3.1?
Debugging and root cause
Multi file failures where the model has to hold a causal chain before answering.
Architecture and planning
Design docs, migration plans and trade off analysis over long context.
Data analysis
Spreadsheets and logs read in one pass, then summarised with the numbers quoted back.
Agentic coding
Repo wide changes that need plan, edit and verify loops.
08 · Cost levers
Prompt caching and how to cut the bill
Cached input bills at $0.382 per million instead of $0.25, a 28% discount. Keep the system prompt and documents first and the changing question last, so every call hits the cache.
09 · Pricing
DeepSeek V3.1 pricing
Official pricing
List price
$0.25 / 1M in
$0.95 / 1M out
- $0.382 cached input
- 164K context
- text input
Cheapest route
$0.25 / 1M in
$0.95 / 1M out
- 4 routes on OpenRouter
- What Hoodgen routes to first
- Same model, lower bill
Highest route
$0.5 / 1M in
$1.5 / 1M out
- Priority or premium tier
- Best under load
- Use for latency critical calls
Market pricing, every route
| Provider | Tier | Input | Output | Cache read | Uptime 24h |
|---|---|---|---|---|---|
| DeepInfra | standard | $0.25 | $0.95 | $0.13 | 100.0% |
| SiliconFlow | standard | $0.27 | $1 | n/a | 97.3% |
| CoreWeave | standard | $0.5 | $1.5 | $0.2 | 100.0% |
| SambaNova | standard | $0.65 | $1.5 | n/a | 0.0% |
API pricing and cost calculator
| Rate | Per call | Total | Credits |
|---|---|---|---|
| Cheapest route | $0.0050 | $2.49 | 2368 |
| List price | $0.0080 | $4.00 | 3802 |
| Highest route | $0.0123 | $6.13 | 5819 |
10 · Developers
API and developer information
| Model ID | deepseek/deepseek-chat-v3.1 |
| Endpoint | https://openrouter.ai/api/v1/chat/completions |
| Parameters | prompt and max_tokens |
| Streaming | Yes, server sent events |
| Modality string | text->text |
| Max prompt tokens | 163,840 |
| Max completion tokens | 147,456 |
curl https://openrouter.ai/api/v1/chat/completions \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek/deepseek-chat-v3.1",
"messages": [{ "role": "user", "content": "Summarise this document." }],
"max_tokens": 1024
}'11 · Limitations
Model limitations
| Text only output | It cannot generate images, audio or video. Pair it with an image model on Hoodgen when you need visuals. |
| Output cap | 147,456 tokens per request. Long deliverables have to be chunked. |
| Benchmarks not published | Artificial Analysis does not publish an Intelligence Index for this release, so no benchmark score is claimed here. |
| Knowledge cutoff | The provider does not state a fixed cutoff for every release. Verify very recent events against search. |
12 · Comparison
DeepSeek V3.1 vs other Deepseek models
| Model | Context | Input /1M | Output /1M | AA index | Released |
|---|---|---|---|---|---|
| DeepSeek V3.1 | 164K | $0.25 | $0.95 | n/a | 2025 |
| DeepSeek V4 Pro 0423 | 1.05M | $0.9 | $1.5 | n/a | 2026 |
| R1 0528 | 164K | $0.5 | $2.15 | n/a | 2025 |
| V4 Flash Vision Exp | 1.05M | $0.216 | $0.647 | n/a | 2026 |
| DeepSeek V4 Pro 0813 | 1.05M | $0.1 | $8 | n/a | 2026 |
| DeepSeek V4.1 Flash | 1.05M | $0.051 | $0.567 | n/a | Sep 2026 |
13 · Privacy
Privacy and data policy
- On Hoodgen, prompts and attachments answer that request and stay in your conversation history so you can reopen it. They are not used to train models.
- Requests route through OpenRouter to Deepseek and are covered by that provider's API terms.
- Do not paste secrets or personal data you would not send to any hosted model.
14 · FAQ
DeepSeek V3.1 questions, answered
DeepSeek V3.1 is a reasoning model from Deepseek, released 21 August 2025. DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates.
✦
Start chatting with DeepSeek V3.1
Free to try on Hoodgen AI, with a live credit estimate before every send. No API key, no setup.