Qwen3.5 397B A17B vs DeepSeek V4 Pro 0813
At a Glance
| Compare |  DeepSeek V4 Pro 0813DeepSeek | |
|---|---|---|
| Intelligence, Cost, and Efficiency | ||
| IntelligenceHigher is better · MM Intelligence v2.5 | UnrankedNot in the 46-model eligible cohort | #20 of 4665.0 score · 3/3 sources · complete |
| CostLower is better · Published-token output estimate | UnrankedNot in the 44-model eligible cohort | #20 of 44$0.094 per LiveBench case |
| EfficiencyHigher is better · MM Efficiency v1.5 | UnrankedNot in the 38-model eligible cohort | #14 of 3857.4 score · 3/3 sources · complete |
| Pricing and Limits | ||
| Input priceFrom · USD / 1M tokens | $0.45Deepinfra ↗ · Sep 22, 2026 | $1.30Deepinfra ↗ · Sep 22, 2026 |
| Output priceFrom · USD / 1M tokens | $3.00Deepinfra ↗ · Sep 22, 2026 | $2.60Deepinfra ↗ · Sep 22, 2026 |
| Context windowMaximum documented tokens | 262K | 1,049K |
| Model facts checked | Aug 28, 2026View model evidence → | Sep 3, 2026View model evidence → |
Token prices are the lowest available sourced USD rates; input and output may use different providers. Cost ranking estimates output spend on LiveBench, not a full request bill. Ranking methodology →
Available Benchmarks
| Benchmark | Qwen3.5-397B-A17B | DeepSeek-V4-Pro-0813 |
|---|---|---|
| LMArena Text Arenatext-2026-09-13-d25aabda0010 · arena_rating · statistical tie | 1,438.26100% of row best · rating · qwen3.5-397b-a17b; 95% CI [1434.97391785, 1441.55191176]; votes 77007; rank 71 | 1,443.95100% of row best · rating · deepseek-v4-pro-high-20260813; 95% CI [1437.16050212, 1450.73220317]; votes 9008; rank 58 |
| Overall ResultCounted from the protocol-matched rows above · 1 tie | 0 benchmark winsNo overall winner | 0 benchmark winsNo overall winner |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Side-by-Side Facts
| Field | Qwen3.5-397B-A17B | DeepSeek-V4-Pro-0813 |
|---|---|---|
| Developer | Qwen | DeepSeek |
| Family | Qwen3 5 397b A17b | Deepseek V4 Pro |
| Model | Qwen3.5-397B-A17B | DeepSeek-V4-Pro-0813 |
| Version | Qwen3.5-397B-A17B | DeepSeek-V4-Pro-0813 |
| Lifecycle | active | active |
| Released | 2026-02-15 | 2026-08-13 |
| Knowledge cutoff | Unknown | Unknown |
| Input modalities | Text, Image | Text |
| Output modalities | Text | Text |
| Context window | 262K | 1,049K |
| Total parameters | 403.4B | 1.6T |
| Active parameters | 17B | 49B |
| License | apache-2.0 | mit |
| Open weights | Yes | Yes |
| API available | Yes | Yes |
| Self-hostable | Yes | Yes |
| Provider access | Deepinfra (Standard), Hugging Face (Standard), Openrouter (Standard), Together Ai (Standard) | Deepinfra (Standard), Fireworks Ai (Standard), Together Ai (Standard) |
| Capabilities | chat, generation, reasoning, tools | agents, chat, coding, generation, reasoning, tools |
Qwen3.5 397B A17B Capabilities
DeepSeek V4 Pro 0813 Capabilities
Related Comparisons
All text comparisons →| A | Pair | B | Context |
|---|---|---|---|
| vs |  GPT-6 AstraOpenAI | cross-developer peersimage, text | |
| vs | Claude Fable 5.1Anthropic | cross-developer peersimage, text | |
| vs |  Gemini 3.1 ProGoogle DeepMind | cross-developer peerstext | |
| vs | DeepSeek-V4.1-FlashDeepSeek | cross-developer peersimage, text | |
| vs | Grok 4.6xAI | cross-developer peersimage, text | |
| vs |  Qwen3.8-MaxQwen | family variantsimage, text | |
| vs |  Kimi-K3Moonshot AI | cross-developer peersimage, text | |
 MiniMax-M3MiniMax | vs | cross-developer peersimage, text | |
| vs |  GLM-5.3Z.ai | cross-developer peerstext | |
| vs |  Hy4 previewTencent | cross-developer peerstext | |
| vs | Seed 2.1 ProByteDance Seed | cross-developer peersimage, text | |
| vs |  Mistral Large 3Mistral AI | cross-developer peersimage, text |
Primary Evidence
Sources and Freshness
Questions
Qwen3.5 397B A17B vs DeepSeek V4 Pro 0813 FAQs
Is Qwen3.5 397B A17B or DeepSeek V4 Pro 0813 better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both Qwen3.5 397B A17B and DeepSeek V4 Pro 0813, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
Qwen3.5 397B A17B is $0.45 and DeepSeek V4 Pro 0813 is $1.30 per million tokens, so Qwen3.5 397B A17B is cheaper on this metric. Qwen3.5 397B A17B is $3.00 and DeepSeek V4 Pro 0813 is $2.60 per million tokens, so DeepSeek V4 Pro 0813 is cheaper on this metric.
Which has a larger context window, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
DeepSeek V4 Pro 0813 has the larger sourced context window. Qwen3.5 397B A17B supports 262K and DeepSeek V4 Pro 0813 supports 1,049K.
Which performs better in benchmarks, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
There is no overall benchmark winner: At least two independently verified, protocol-matched benchmarks are required for an overall winner.
Can Qwen3.5 397B A17B or DeepSeek V4 Pro 0813 be self-hosted?+
Both models have the same recorded self-hosting status: supported. Qwen3.5 397B A17B is open weight; DeepSeek V4 Pro 0813 is open weight.
Can Qwen3.5 397B A17B and DeepSeek V4 Pro 0813 understand images?+
Qwen3.5 397B A17B is documented with image input; DeepSeek V4 Pro 0813 is not documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
Neither has a larger sourced maximum output. Qwen3.5 397B A17B is — and DeepSeek V4 Pro 0813 is —.
Do Qwen3.5 397B A17B and DeepSeek V4 Pro 0813 support reasoning and tool use?+
Qwen3.5 397B A17B: reasoning, tool calling, and image input. DeepSeek V4 Pro 0813: reasoning and tool calling. Feature support does not establish relative quality.
Which is available from more inference providers, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
Qwen3.5 397B A17B has 4 sourced provider routes; DeepSeek V4 Pro 0813 has 3, so Qwen3.5 397B A17B has broader tracked availability.
Which offers better value, Qwen3.5 397B A17B or DeepSeek V4 Pro 0813?+
There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.