Model comparison
DeepSeek V3.2 (Thinking) vs Kimi K2.5
Head-to-head evidence from 1 shared benchmark result across 1 category. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: DeepSeek V3.2 (Thinking) #65 (Estimated); Kimi K2.5 #54 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. DeepSeek V3.2 (Thinking) and Kimi K2.5 share 1 comparable benchmark result. 0 of 8 categories are comparable. 1 result is unique to DeepSeek V3.2 (Thinking); 62 to Kimi K2.5.
Updated July 20, 2026- Shared results
- 1
- DeepSeek V3.2 (Thinking) only
- 1
- Kimi K2.5 only
- 62
- Comparable categories
- 0 / 8
Benchmark data for DeepSeek V3.2 (Thinking) and Kimi K2.5 is coming soon on BenchLM.
Confidence note. This is a partial-evidence comparison with 1 shared benchmark result across 1 evidence category; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.
Kimi K2.5 is priced at $0.60 input / $3.00 output per 1M tokens, versus $0.55 input / $2.19 output per 1M tokens for DeepSeek V3.2 (Thinking). Kimi K2.5 has the larger context window at 256K, compared with 128K for DeepSeek V3.2 (Thinking).
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | DeepSeek V3.2 (Thinking) | Δ | Kimi K2.5 |
|---|---|---|---|
| Agentic | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.555.0 |
| Coding | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.559.4 |
| Reasoning | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.561.0 |
| Knowledge | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.556.9 |
| Math | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.560.6 |
| Multilingual | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.582.3 |
| Multimodal | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.578.5 |
| Inst. Following | DeepSeek V3.2 (Thinking)Not measured | MarginNo overlap | Kimi K2.593.9 |
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | DeepSeek V3.2 (Thinking) | Kimi K2.5 | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | DeepSeek V3.2 (Thinking)$0.55 input / $2.19 output | Kimi K2.5$0.6 input / $3 output | DeepSeek V3.2 (Thinking) has the lower combined listed price. |
| Generation speedtokens per second | DeepSeek V3.2 (Thinking)Not available | Kimi K2.545 tok/s | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | DeepSeek V3.2 (Thinking)Not available | Kimi K2.52.38 s | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | DeepSeek V3.2 (Thinking)128K | Kimi K2.5256K | Kimi K2.5 lists the larger context window. |
Benchmark Deep Dive
Agentic19 benchmarks
| Benchmark | DeepSeek V3.2 (Thinking) | Kimi K2.5 | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | — | 50.8% | Not comparable |
| BrowseCompSource | — | 60.6% | Not comparable |
| Claw-EvalSource | — | 52.3% | Not comparable |
| QwenClawBenchSource | — | 54.3% | Not comparable |
| τ³-bench resultsSource | — | 65.7% | Not comparable |
| DeepSearchQASource | — | 77.1% | Not comparable |
| DeepPlanningSource | — | 14.4% | Not comparable |
| ToolathlonSource | — | 27.8% | Not comparable |
| MCP AtlasSource | — | 29.5% | Not comparable |
| MCP-TasksSource | — | 59.1% | Not comparable |
| WideResearchSource | — | 72.7% | Not comparable |
| τ²-bench resultsSource | — | 95.9% | Not comparable |
| APEX-Agents-AASource | — | 11.5% | Not comparable |
| Gert LabsSource | — | 45.88% | Not comparable |
| ResearchClawBenchSource | — | 14.0% | Not comparable |
| JobBenchSource | — | 8.7% | Not comparable |
| AA Agentic IndexSource | — | 21.7% | Not comparable |
| GDPval-AASource | — | 25.4% | Not comparable |
| GDPval-AASource | — | 1009 | Not comparable |
Coding11 benchmarks
| Benchmark | DeepSeek V3.2 (Thinking) | Kimi K2.5 | Result |
|---|---|---|---|
| Vibe Code BenchSource | 5.11% | — | Not comparable |
| SWE-bench VerifiedSource | — | 76.8% | Not comparable |
| SWE-bench Verified*Source | — | 70.8% | Not comparable |
| LiveCodeBench v6Source | — | 85.0% | Not comparable |
| SWE-bench ProSource | — | 50.7% | Not comparable |
| SWE MultilingualSource | — | 73% | Not comparable |
| SWE-RebenchSource | — | 58.5% | Not comparable |
| React Native EvalsSource | — | 77.2% | Not comparable |
| SciCodeSource | — | 48.7% | Not comparable |
| AA-SciCodeSource | — | 49.0% | Not comparable |
| AA Coding IndexSource | — | 46.8% | Not comparable |
Reasoning3 benchmarks
Knowledge12 benchmarks
| Benchmark | DeepSeek V3.2 (Thinking) | Kimi K2.5 | Result |
|---|---|---|---|
| GPQASource | — | 87.6% | Not comparable |
| GPQA-DSource | — | 87.6% | Not comparable |
| SuperGPQASource | — | 69.2% | Not comparable |
| MMLU-ProSource | — | 87.1% | Not comparable |
| MMLU-Pro (Arcee)Source | — | 87.1% | Not comparable |
| HLESource | — | 30.1% | Not comparable |
| Artificial Analysis Intelligence IndexSource | — | 35.4% | Not comparable |
| AA-GPQA DiamondSource | — | 87.9% | Not comparable |
| AA-HLESource | — | 29.4% | Not comparable |
| AA-Omniscience IndexSource | — | -8.1% | Not comparable |
| AA-Omniscience AccuracySource | — | 34.3% | Not comparable |
| AA-Omniscience Hallucination RateSource | — | 64.6% | Not comparable |
Math9 benchmarks
| Benchmark | DeepSeek V3.2 (Thinking) | Kimi K2.5 | Result |
|---|---|---|---|
| AIME 2025Source | — | 96.1% | Not comparable |
| AIME26Source | — | 95.8% | Not comparable |
| AIME25 (Arcee)Source | — | 96.3% | Not comparable |
| HMMT Feb 2025Source | — | 95.4% | Not comparable |
| HMMT Nov 2025Source | — | 91.1% | Not comparable |
| HMMT Feb 2026Source | — | 87.1% | Not comparable |
| MMAnswerBenchSource | — | 81.8% | Not comparable |
| FrontierMath v2 (Tiers 1-3)Source | — | 27.900% | Not comparable |
| FrontierMath v2 (Tier 4)Source | — | 4.200% | Not comparable |
Multilingual2 benchmarks
Multimodal6 benchmarks
Frequently Asked Questions (3)
Can I compare DeepSeek V3.2 (Thinking) and Kimi K2.5 on BenchLM yet?
Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.
Why does this comparison show “coming soon”?
BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.
What data is available for DeepSeek V3.2 (Thinking) and Kimi K2.5 today?
DeepSeek V3.2 (Thinking): $0.55 input / $2.19 output per 1M tokens Kimi K2.5: $0.60 input / $3.00 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Related Comparisons
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.