DeepSeek-V4.1-FlashvsKimi K3
Across 15 shared benchmarks, DeepSeek-V4.1-Flash leads overall: DeepSeek-V4.1-Flash wins 13, Kimi K3 wins 1, with 1 ties and an average score difference of +6.35.
DeepSeek-V4.1-Flash
DeepSeek-AI · 2026-09-10 · Multimodal model
Kimi K3
Moonshot AI · 2026-07-16 · Reasoning model
DeepSeek-V4.1-Flash13 wins(87%)Ties1(7%)1 winKimi K3
Benchmark scores
Grouped by capability, sorted by largest gap within each. 15 shared benchmarks.
AI Agent - Tool Usage
DeepSeek-V4.1-Flash 4/4| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| Terminal-Bench 4.0 | 31.207 / 16Max (With Tools) | 12.6013 / 16Max (With Tools) | +18.60 |
| Terminal-Bench 3.0 | 304 / 11Max (With Tools) | 17.708 / 11Max (With Tools) | +12.30 |
| CyberGym | 88.101 / 8Max (With Tools) | 805 / 8Max (With Tools) | +8.10 |
| Terminal-Bench 2.1 | 90.601 / 53Max (With Tools) | 88.307 / 53Max (With Tools) | +2.30 |
Coding and Software Engineer
DeepSeek-V4.1-Flash 3/3| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| NL2Repo-Bench | 65.402 / 16Max (With Tools) | 586 / 16Max (With Tools) | +7.40 |
| DeepSWE | 74.202 / 38Max (With Tools) | 67.5011 / 38Max (With Tools) | +6.70 |
| Program Bench | 20.308 / 11Max (With Tools) | 17.5010 / 11Max (With Tools) | +2.80 |
Multimodal Understanding
DeepSeek-V4.1-Flash 3/3| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| Chartography | 78.904 / 8Max (With Tools) | 68.106 / 8Max (With Tools) | +10.80 |
| ZeroBench Main | 493 / 6Max (With Tools) | 414 / 6Max (With Tools) | +8 |
| BabyVision | 89.603 / 9Max (With Tools) | 85.705 / 9Max (With Tools) | +3.90 |
Agent Level Benchmark
DeepSeek-V4.1-Flash 1/1| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| Agents' Last Exam | 31.806 / 19Max (With Tools) | 27.609 / 19Max (With Tools) | +4.20 |
General Evaluation
Kimi K3 1/1| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| GPQA Diamond | 90.9039 / 274Max (No Tools) | 92.9023 / 274Max (No Tools) | -2 |
General Knowledge
DeepSeek-V4.1-Flash 1/1| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| HLE | 63.903 / 197Max (With Tools) | 59.808 / 197Max (With Tools) | +4.10 |
Math and Reasoning
Even 1/1| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| MathArena Apex | 65.601 / 3Max (No Tools) | 65.601 / 3Max (No Tools) | — |
Productivity Knowledge
DeepSeek-V4.1-Flash 1/1| Benchmark | DeepSeek-V4.1-Flash | Kimi K3 | Diff |
|---|---|---|---|
| AutomationBench | 54.801 / 17Max (With Tools) | 46.707 / 17Max (With Tools) | +8.10 |
Specs
| Field | DeepSeek-V4.1-Flash | Kimi K3 |
|---|---|---|
| Publisher | DeepSeek-AI | Moonshot AI |
| Release date | 2026-09-10 | 2026-07-16 |
| Model type | Multimodal model | Reasoning model |
| Architecture | MoE | MoE |
| Parameters | 552B | 2.8T |
| Context length | 1M | 1M |
| Max output | 384K | 1M |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | DeepSeek-V4.1-Flash | Kimi K3 |
|---|---|---|
| Text input | ¥1 / 1M tokens | ¥20 / 1M tokens |
| Text output | ¥4 / 1M tokens | ¥100 / 1M tokens |
| Cache read | ¥0.02 / 1M tokens | ¥2 / 1M tokens |
Summary
- DeepSeek-V4.1-Flashleads in:AI Agent - Tool Usage (4/4), Coding and Software Engineer (3/3), Multimodal Understanding (3/3), Agent Level Benchmark (1/1), General Knowledge (1/1), Productivity Knowledge (1/1)
- Kimi K3leads in:General Evaluation (1/1)
- Tied in:Math and Reasoning
On average across the 15 shared benchmarks, DeepSeek-V4.1-Flash scores 6.35 higher.
Largest single-benchmark gap: Terminal-Bench 4.0 — DeepSeek-V4.1-Flash 31.20 vs Kimi K3 12.60 (+18.60).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.