DeepSeek V4 Pro 0813 vs Kimi K3
Why this pair: Current open-weight agentic flagship models from DeepSeek and Moonshot AI
Model Markets Rankings
Intelligence, Cost, and Efficiency
| Ranking | DeepSeek-V4-Pro-0813 | Kimi-K3 |
|---|---|---|
| CostLower is better · Published-token output estimate | UnrankedNot in the 36-model eligible cohort | #24 of 36$0.194 per LiveBench case |
Ranks come from the current complete eligible cohorts. Green highlights appear only when both models are ranked in the same metric. Missing required inputs remain unranked, and the three dimensions are not collapsed into an overall winner.
Benchmark Performance
Available Benchmarks
| Benchmark | DeepSeek-V4-Pro-0813 | Kimi-K3 |
|---|---|---|
| ARC-AGI-1verified-v1-2aa739d957fc · verified_score · leader | 90.5096% of row best · percent · DeepSeek V4 Pro 0813 (Low) | 94.50100% of row best · percent · Kimi K3 (Max) |
| ARC-AGI-2verified-v2-b5cb5ec6e8c7 · verified_score · leader | 61.25100% of row best · percent · DeepSeek V4 Pro 0813 (Max) | 60.4299% of row best · percent · Kimi K3 (Max) |
| ToneBench2026-08-28-10-task-cd9819ab6e4d · overall_score · leader | 84.0495% of row best · points · DeepSeek V4 Pro 0813 (max) · 18,922 output tokens / case | 88.19100% of row best · points · Kimi K3 · 15,023 output tokens / case |
| Overall ResultCounted from the protocol-matched rows above | 1 benchmark win | 2 benchmark winsOverall lead |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Technical Differences
Side-by-Side Facts
| Field | DeepSeek-V4-Pro-0813 | Kimi-K3 |
|---|---|---|
| Developer | DeepSeek | Moonshot AI |
| Family | Deepseek V4 Pro | Kimi K3 |
| Model | DeepSeek-V4-Pro-0813 | Kimi-K3 |
| Version | DeepSeek-V4-Pro-0813 | Kimi-K3 |
| Lifecycle | active | active |
| Released | 2026-08-13 | 2026-07-16 |
| Knowledge cutoff | Unknown | Unknown |
| Input modalities | Text | Text, Image |
| Output modalities | Text | Text |
| Context window | 1,049K | 1,049K |
| Total parameters | 1.6T | 2.8T |
| Active parameters | 49B | 104B |
| License | mit | other |
| Open weights | Yes | Yes |
| API available | Yes | Yes |
| Self-hostable | Yes | Yes |
| Provider access | Deepinfra (Standard), Fireworks Ai (Standard), Together Ai (Standard) | Deepinfra (Standard), Fireworks Ai (Standard), Hugging Face (Standard), Openrouter (Standard), Together Ai (Standard) |
| Capabilities | agents, chat, coding, generation, reasoning, tools | chat, generation, reasoning |
17 comparable fields · 11 material differences · Editorially curated pair passes the primary-source comparison gate
DeepSeek V4 Pro 0813 Capabilities
Kimi K3 Capabilities
Internal Comparison Graph
Related Comparisons
| A | Pair | B | Context |
|---|---|---|---|
DeepSeek-V4-Pro-0813DeepSeek | vs |  GPT-6 AstraOpenAI | cross-developer peerstext |
Claude Fable 5.1Anthropic | vs | DeepSeek-V4-Pro-0813DeepSeek | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs |  Gemini 3.1 ProGoogle DeepMind | cross-developer peerstext |
DeepSeek-V4-ProDeepSeek | vs | DeepSeek-V4-Pro-0813DeepSeek | family variantstext |
DeepSeek-V4-Pro-0813DeepSeek | vs | Grok 4.6xAI | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs |  Qwen3.8-MaxQwen | cross-developer peerstext |
 MiniMax-M3MiniMax | vs | DeepSeek-V4-Pro-0813DeepSeek | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs |  GLM-5.3Z.ai | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs |  Hy4 previewTencent | cross-developer peerstext |
Seed 2.1 ProByteDance Seed | vs | DeepSeek-V4-Pro-0813DeepSeek | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs |  Mistral Large 3Mistral AI | cross-developer peerstext |
DeepSeek-V4-Pro-0813DeepSeek | vs | cross-developer peerstext |
Primary Evidence
Sources and Freshness
Questions
DeepSeek V4 Pro 0813 vs Kimi K3 FAQs
Is DeepSeek V4 Pro 0813 or Kimi K3 better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both DeepSeek V4 Pro 0813 and Kimi K3, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, DeepSeek V4 Pro 0813 or Kimi K3?+
DeepSeek V4 Pro 0813 is $1.30 and Kimi K3 is $2.85 per million tokens, so DeepSeek V4 Pro 0813 is cheaper on this metric. DeepSeek V4 Pro 0813 is $2.60 and Kimi K3 is $14.25 per million tokens, so DeepSeek V4 Pro 0813 is cheaper on this metric.
Which has a larger context window, DeepSeek V4 Pro 0813 or Kimi K3?+
Neither model has a larger sourced context window in this comparison. DeepSeek V4 Pro 0813 is 1,049K and Kimi K3 is 1,049K.
Which performs better in benchmarks, DeepSeek V4 Pro 0813 or Kimi K3?+
Kimi K3 leads the current overall benchmark count. The result uses 3 protocol-matched benchmarks from 2 publishers; it is not a universal quality score.
Can DeepSeek V4 Pro 0813 or Kimi K3 be self-hosted?+
Both models have the same recorded self-hosting status: supported. DeepSeek V4 Pro 0813 is open weight; Kimi K3 is open weight.
Can DeepSeek V4 Pro 0813 and Kimi K3 understand images?+
DeepSeek V4 Pro 0813 is not documented with image input; Kimi K3 is documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, DeepSeek V4 Pro 0813 or Kimi K3?+
Neither has a larger sourced maximum output. DeepSeek V4 Pro 0813 is — and Kimi K3 is —.
Do DeepSeek V4 Pro 0813 and Kimi K3 support reasoning and tool use?+
DeepSeek V4 Pro 0813: reasoning and tool calling. Kimi K3: reasoning and image input. Feature support does not establish relative quality.
Which is available from more inference providers, DeepSeek V4 Pro 0813 or Kimi K3?+
DeepSeek V4 Pro 0813 has 3 sourced provider routes; Kimi K3 has 5, so Kimi K3 has broader tracked availability.
Which offers better value, DeepSeek V4 Pro 0813 or Kimi K3?+
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.