
DeepSeek: DeepSeek V4.1 Flash
deepseek/deepseek-v4.1-flash
DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on output from a 552B-parameter backbone, an asymmetric split that keeps per-token compute low relative to the model's total size. Image understanding is native to the architecture, with visual and text embeddings trained jointly from the start of pre-training rather than added afterward as in the earlier experimental V4 Flash Vision ExpOpens in new tab.
It is suited for coding, terminal, and computer-use agents, along with long-horizon tasks that must run to completion across many steps and long-context analysis. Compressed KV caching cuts cache memory to roughly a quarter of the previous Flash generation, significantly reducing costs on agentic workloads. DeepSeek positions it as the cost-efficient tier of the V4.1 family and reports that it exceeds V4 ProOpens in new tab on performance, speed, and task completion time.
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
| $0.1950 | $0.660 | $0.00300 | 0.37s | 160 tok/s | ||
| $0.1030 | $0.600 | $0.00800 | 0.25s | 181 tok/s | ||
| $0.0510 | $0.567 | $0.01800 | 0.37s | 173 tok/s | ||
| $0.1030 | $0.600 | $0.00800 | 0.25s | 166 tok/s | ||
| $0.1200 | $0.480 | $0.00240 | 0.73s | 123 tok/s | ||
| $0.1900 | $0.650 | $0.01200 | 0.33s | 182 tok/s | ||
| $0.1300 | $0.650 | $0.00200 | 0.63s | 124 tok/s | ||
| $0.1200 | $0.500 | $0.00300 | 0.94s | 115 tok/s | ||
| $0.1800 | $0.720 | $0.00360 | 0.76s | 135 tok/s | ||
| $0.20$0.1400 | $0.60$0.420 | $0.006$0.00420 | 0.78s | 96 tok/s | ||
| $0.2250 | $0.900 | $0.00600 | 0.72s | 149 tok/s | ||
| $0.3000 | $1.200 | $0.00700 | 0.61s | 214 tok/s | ||
| $0.3000 | $1.200 | $0.00600 | 0.22s | 261 tok/s | ||
60% off | $0.30$0.1200 | $1.20$0.480 | $0.006$0.00240 | 0.92s | 106 tok/s | |
| $0.2900 | $1.200 | $0.00700 | 0.55s | 174 tok/s | ||
| $0.3000 | $1.200 | $0.00600 | 0.49s | 201 tok/s | ||
| $0.3000 | $1.200 | $0.00750 | 0.58s | 176 tok/s | ||
| $0.0900 | $1.200 | $0.00500 | 1.01s | 142 tok/s | ||
| $0.3000 | $1.200 | $0.03000 | 0.39s | 166 tok/s | ||
51% off | $0.30$0.1470 | $1.20$0.588 | $0.006$0.00294 | 1.30s | 86 tok/s | |
| $0.2100 | $0.840 | $0.00420 | 0.65s | 114 tok/s | ||
| $0.30$0.2100 | $0.90$0.630 | $0.009$0.00630 | 0.89s | 84 tok/s | ||
| $0.3000 | $1.200 | $0.00600 | 0.39s | 159 tok/s | ||
20% off | $0.15$0.1200 | $0.80$0.640 | $0.01$0.00800 | 0.28s | 76 tok/s | |
| $0.2600 | $1.170 | $0.00580 | 0.70s | 147 tok/s | ||
| $0.1350 | $0.540 | $0.01350 | 0.71s | 69 tok/s | ||
| $0.4500 | $1.800 | $0.00900 | 0.45s | 207 tok/s | ||
Fast | $0.4500 | $1.800 | $0.00900 | 0.40s | 199 tok/s | |
| $0.2125 | $0.750 | $0.05000 | 0.88s | 78 tok/s | ||
| $0.3000 | $1.200 | $0.00600 | 1.58s | 126 tok/s | ||
40% off | $0.30$0.1800 | $1.20$0.720 | $0.006$0.00360 | 1.68s | 47 tok/s | |
| $0.3000 | $1.200 | $0.03000 | 0.69s | 71 tok/s | ||
| $0.3000 | $1.200 | $0.00700 | 0.29s | 195 tok/s | ||
| $0.1000 | $0.900 | $0.03000 | 1.44s | 77 tok/s | ||
| $0.3000 | $1.200 | $0.00600 | 1.23s | 53 tok/s | ||
Not used in Standard routing:Why these endpoints are not used | ||||||
Ultrafast | $3.0000 | $12.000 | $0.06000 | -- | -- | |
| $0.16$0.1120 | $0.48$0.336 | $0.0048$0.00336 | 1.87s | 77 tok/s | ||
Input $0.195 / Output $0.66160 tok/s99.88% uptime- Input $0.103 / Output $0.60181 tok/s100.00% uptime
- Input $0.051 / Output $0.567173 tok/s100.00% uptime
- Input $0.103 / Output $0.60166 tok/s100.00% uptime
- Input $0.12 / Output $0.48123 tok/s100.00% uptime
- Input $0.19 / Output $0.65182 tok/s100.00% uptime
- Input $0.13 / Output $0.65124 tok/s96.94% uptime
- Input $0.12 / Output $0.50115 tok/s100.00% uptime
- Input $0.18 / Output $0.72135 tok/s99.57% uptime
Input $0.14 / Output $0.4230% off96 tok/s100.00% uptime- Input $0.225 / Output $0.90149 tok/s100.00% uptime
- Input $0.30 / Output $1.20214 tok/s100.00% uptime
- Input $0.30 / Output $1.20261 tok/s100.00% uptime
- Input $0.12 / Output $0.4860% off106 tok/s100.00% uptime
- Input $0.29 / Output $1.20174 tok/s100.00% uptime
- Input $0.30 / Output $1.20201 tok/s100.00% uptime
- Input $0.30 / Output $1.20176 tok/s100.00% uptime
- Input $0.09 / Output $1.20142 tok/s99.83% uptime
- Input $0.30 / Output $1.20166 tok/s100.00% uptime
- Input $0.147 / Output $0.58851% off86 tok/s100.00% uptime
Input $0.21 / Output $0.84114 tok/s99.57% uptime
Input $0.21 / Output $0.6384 tok/s100.00% uptimePriority+11 more label
Input $0.30 / Output $1.20159 tok/s99.85% uptime- Input $0.12 / Output $0.6420% off76 tok/s99.19% uptime
- Input $0.26 / Output $1.17147 tok/s99.97% uptime
- Input $0.135 / Output $0.5469 tok/s100.00% uptime
Input $0.45 / Output $1.80207 tok/s100.00% uptime- Input $0.45 / Output $1.80199 tok/s100.00% uptimeFast
- Input $0.2125 / Output $0.7578 tok/s99.73% uptime
- Input $0.30 / Output $1.20126 tok/s99.88% uptime
- Input $0.18 / Output $0.7240% off47 tok/s99.91% uptime
- Input $0.30 / Output $1.2071 tok/s95.16% uptime
- Input $0.30 / Output $1.20195 tok/s100.00% uptime
- Input $0.10 / Output $0.9077 tok/s100.00% uptime
- Input $0.30 / Output $1.2053 tok/s100.00% uptime
- Not used in Standard routing:Why these endpoints are not used
- Input $3.00 / Output $12.00-- tok/s100.00% uptimeUltrafast
Input $0.112 / Output $0.33677 tok/s98.89% uptimeFlex+11 more label
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
DeepSeek V4.1 Flash — Price history
| $0.00605 | $0.587 | 97.8% | 2.6% | ||
| $0.00734 | $1.200 | 97.4% | 10.4% | ||
| $0.01211 | $0.424 | 90.4% | 4.7% | ||
| $0.01384 | $0.500 | 90.7% | 1.7% | ||
| $0.01554 | $0.600 | 92.1% | 5.0% | ||
| $0.01607 | $0.720 | 92.9% | 1.7% | ||
| $0.01715 | $0.421 | 90.5% | 11.5% | ||
| $0.01782 | $0.600 | 89.7% | 3.5% | ||
| $0.01948 | $0.660 | 91.4% | 4.9% | ||
| $0.02066 | $0.900 | 93.3% | 3.2% | ||
| $0.02078 | $0.567 | 91.5% | 12.9% | ||
| $0.02338 | $0.485 | 82.3% | 1.0% | ||
| $0.02589 | $0.840 | 89.5% | 1.2% | ||
| $0.02686 | $0.650 | 80.6% | 3.7% | ||
| $0.02811 | $0.630 | 89.3% | 0.5% | ||
| $0.02846 | $1.200 | 92.4% | 2.2% | ||
| $0.02964 | $1.170 | 90.6% | 0.5% | ||
| $0.03004 | $1.200 | 92.1% | 2.4% | ||
| $0.03043 | $1.200 | 91.7% | 1.8% | ||
| $0.03096 | $0.638 | 79.4% | 0.1% | ||
| $0.03293 | $0.650 | 88.2% | 1.4% | ||
| $0.03727 | $0.726 | 81.0% | 0.0% | ||
| $0.03986 | $1.800 | 93.0% | 1.4% | ||
| $0.04015 | $1.200 | 88.8% | 0.9% | ||
| $0.04072 | $0.898 | 84.7% | 0.1% | ||
| $0.04231 | $1.199 | 87.9% | 0.1% | ||
| $0.04346 | $1.200 | 87.3% | 0.2% | ||
| $0.05474 | $1.199 | 90.8% | 0.0% | ||
| $0.05558 | $1.200 | 83.1% | 19.9% | ||
| $0.05996 | $0.536 | 61.7% | 0.2% | ||
| $0.07446 | $1.200 | 76.7% | 0.1% | ||
| $0.09022 | $0.748 | 74.9% | 0.0% | ||
Fast | $0.12105 | $1.800 | 74.6% | 0.1% | |
| $0.13483 | $1.198 | 56.2% | 0.1% | ||
| $0.17963 | $1.200 | 44.5% | 0.1% | ||
Not used in Standard routing:Why these endpoints are not used | |||||
Ultrafast | $1.47151 | $12.000 | 52.0% | 0.0% | |
| $0.01779 | $0.336 | 86.7% | 0.2% | ||
- Effective in $0.00605 / Effective out $0.587Token share2.6%
- Effective in $0.00734 / Effective out $1.200Token share10.4%
- Effective in $0.01211 / Effective out $0.424Token share4.7%
- Effective in $0.01384 / Effective out $0.500Token share1.7%
- Effective in $0.01554 / Effective out $0.600Token share5.0%
- Effective in $0.01607 / Effective out $0.720Token share1.7%
- Effective in $0.01715 / Effective out $0.421Token share11.5%
- Effective in $0.01782 / Effective out $0.600Token share3.5%
- Effective in $0.01948 / Effective out $0.660Token share4.9%
- Effective in $0.02066 / Effective out $0.900Token share3.2%
- Effective in $0.02078 / Effective out $0.567Token share12.9%
- Effective in $0.02338 / Effective out $0.485Token share1.0%
- Effective in $0.02589 / Effective out $0.840Token share1.2%
- Effective in $0.02686 / Effective out $0.650Token share3.7%
- Effective in $0.02811 / Effective out $0.630PriorityToken share0.5%
- Effective in $0.02846 / Effective out $1.200Token share2.2%
- Effective in $0.02964 / Effective out $1.170Token share0.5%
- Effective in $0.03004 / Effective out $1.200Token share2.4%
- Effective in $0.03043 / Effective out $1.200Token share1.8%
- Effective in $0.03096 / Effective out $0.638Token share0.1%
- Effective in $0.03293 / Effective out $0.650Token share1.4%
- Effective in $0.03727 / Effective out $0.726Token share0.0%
- Effective in $0.03986 / Effective out $1.800Token share1.4%
- Effective in $0.04015 / Effective out $1.200Token share0.9%
- Effective in $0.04072 / Effective out $0.898Token share0.1%
- Effective in $0.04231 / Effective out $1.199Token share0.1%
- Effective in $0.04346 / Effective out $1.200Token share0.2%
- Effective in $0.05474 / Effective out $1.199Token share0.0%
- Effective in $0.05558 / Effective out $1.200Token share19.9%
- Effective in $0.05996 / Effective out $0.536Token share0.2%
- Effective in $0.07446 / Effective out $1.200Token share0.1%
- Effective in $0.09022 / Effective out $0.748Token share0.0%
- Effective in $0.12105 / Effective out $1.800FastToken share0.1%
- Effective in $0.13483 / Effective out $1.198Token share0.1%
- Effective in $0.17963 / Effective out $1.200Token share0.1%
- Not used in Standard routing:Why these endpoints are not used
- Effective in $1.47151 / Effective out $12.000UltrafastToken share0.0%
- Effective in $0.01779 / Effective out $0.336FlexToken share0.2%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
P50, best across providers
P50, best provider
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Availability over the last 3 days
Last 72 hoursAvailability over the last 24 hours
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
| Source | Benchmark | Score |
|---|---|---|
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) HLE | 10.8% |
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) AA-LCR | 55.3% |
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) GDPval-AA | 42.4% |
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) CritPt | 0.3% |
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) SciCode | 35.5% |
| Artificial Analysis | DeepSeek V4.1 Flash (Non-reasoning) Terminal-Bench 4.0 | 5.6% |
| Vals AI | DeepSeek V4.1 Flash Access to Justice Benchmark | 72.9% |
| Vals AI | DeepSeek V4.1 Flash BioMysteryBench | 67.8% |
| Vals AI | DeepSeek V4.1 Flash Code Migration v1.1 | 58.9% |
| Vals AI | DeepSeek V4.1 Flash CyberBench v1.1 | 73.7% |
| Vals AI | DeepSeek V4.1 Flash EMB v1.1 | 67.8% |
| Vals AI | DeepSeek V4.1 Flash Finance Agent v2.1 | 65.4% |
| Vals AI | DeepSeek V4.1 Flash Vals Finance Index | 57.1% |
| Vals AI | DeepSeek V4.1 Flash Financial Coaching Bench | 72.8% |
| Vals AI | DeepSeek V4.1 Flash Harvey's Legal Agent Benchmark v1 | 6.7% |
| Vals AI | DeepSeek V4.1 Flash Immigration | 65.6% |
| Vals AI | DeepSeek V4.1 Flash Vals Inference Provider Index | 66.6% |
| Vals AI | DeepSeek V4.1 Flash IOI | 40.3% |
| Vals AI | DeepSeek V4.1 Flash LegalBench | 83.3% |
| Vals AI | DeepSeek V4.1 Flash Legal Research Bench | 41.3% |
| Vals AI | DeepSeek V4.1 Flash MedCode | 41.2% |
| Vals AI | DeepSeek V4.1 Flash MedScribe | 85.5% |
| Vals AI | DeepSeek V4.1 Flash MysteryMechanism | 21.2% |
| Vals AI | DeepSeek V4.1 Flash ProgramBench | 0.5% |
| Vals AI | DeepSeek V4.1 Flash ProofBench v1.1 | 54.0% |
| Vals AI | DeepSeek V4.1 Flash Public Benefits Bench v1.1 | 64.3% |
| Vals AI | DeepSeek V4.1 Flash SAGE | 85.4% |
| Vals AI | DeepSeek V4.1 Flash Vals Science Index | 29.6% |
| Vals AI | DeepSeek V4.1 Flash SkillsBench | 69.8% |
| Vals AI | DeepSeek V4.1 Flash SRE Bench | 0.8% |
| Vals AI | DeepSeek V4.1 Flash SWE-bench Multimodal | 46.5% |
| Vals AI | DeepSeek V4.1 Flash Tax Agent Bench | 26.6% |
| Vals AI | DeepSeek V4.1 Flash Terminal-Bench 4.0 | 19.7% |
| Vals AI | DeepSeek V4.1 Flash Terminal-Bench-Science | 0.0% |
| Vals AI | DeepSeek V4.1 Flash Vals Index | 51.3% |
| Vals AI | DeepSeek V4.1 Flash Vibe Code Bench 1-100 | 16.4% |
| Vals AI | DeepSeek V4.1 Flash Vibe Code Bench v1.1 | 84.7% |
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
DeepSeek: DeepSeek Flash versions
Not currently available through EU in-region routing. In-region routing docs
Documentation
Guides, cookbooks, blog posts, and tools for building with DeepSeek: DeepSeek V4.1 Flash on OpenRouter.
- Self-Hosted LLM vs API: When Does Self-Hosting Pay OffBlog post
- How to Benchmark LLMsBlog post
- How to Stop an AI Agent Stuck in an Infinite LoopBlog post
- LLM API Rate Limits: How to Fix OpenAI and Claude 429 ErrorsBlog post
- How to Get Structured Output from Any LLM: JSON Schema, Tool Calls, or a Decision ModelBlog post
- Best OCR Model for Scanned Documents, MeasuredBlog post
- Build an AI Code Reviewer with OpenRouterBlog post
- Open-Weight vs Closed-Weight Models: What's the Difference?Blog post
- How to Extract Tables from a PDF with an LLMBlog post
- Does DeepSeek V4 Have Vision?Blog post
Frequently asked questions
DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on output from a 552B-parameter backbone, an asymmetric split that keeps per-token compute low relative to the model's total size.
The DeepSeek V4.1 Flash documentation on OpenRouter includes the Self-Hosted LLM vs API: When Does Self-Hosting Pay Off blog post, the How to Benchmark LLMs blog post, the How to Stop an AI Agent Stuck in an Infinite Loop blog post, the LLM API Rate Limits: How to Fix OpenAI and Claude 429 Errors blog post, the How to Get Structured Output from Any LLM: JSON Schema, Tool Calls, or a Decision Model blog post, the Best OCR Model for Scanned Documents, Measured blog post, the Build an AI Code Reviewer with OpenRouter blog post, the Open-Weight vs Closed-Weight Models: What's the Difference? blog post, the How to Extract Tables from a PDF with an LLM blog post, and the Does DeepSeek V4 Have Vision? blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.
DeepSeek V4.1 Flash costs $0.051/M input tokens and $0.567/M output tokens, with separate rates for Cache Read at $0.018/M tokens.
DeepSeek V4.1 Flash has a 1,048,576 token context window. It supports up to 1,048,576 completion tokens.
Yes. DeepSeek V4.1 Flash accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
DeepSeek V4.1 Flash accepts text and images as input and returns text.
DeepSeek V4.1 Flash is served by 28 providers on OpenRouter: Wafer, Open Inference, Cloudflare, inference.net, DeepInfra, River, NovitaAI, Relace and 20 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
DeepSeek V4.1 Flash was released on September 10, 2026.