# Kimi K2.5 vs Sarvam 30B

> Side-by-side benchmark comparison for Kimi K2.5 and Sarvam 30B across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/kimi-k2-5-vs-sarvam-30b
- Last updated: October 17, 2026

- Shared sourced benchmarks: 5
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

Kimi K2.5 leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests.

## Summary

- Kimi K2.5 has the higher overall point estimate 53.05 to 28.61. Conditional score ranges do not establish rank confidence.
- The clearest category separation is in instruction following, where the averages are 85.8 for Kimi K2.5 and 28.8 for Sarvam 30B.
- The biggest single benchmark swing is SWE-bench Verified in Coding, with scores of 76.8% and 34%.
- Sarvam 30B is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- Kimi K2.5 also has the larger context window at 256K.

## Model Snapshot

| Property | Kimi K2.5 | Sarvam 30B |
|----------|----------|----------|
| Creator | Moonshot AI | Sarvam |
| Type | Open Weight | Open Weight |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 256K | 64K |
| Overall Score | 53.05 | 28.61 |
| Benchmarks Covered | 68 | 9 |
| Pricing (input/output) | $0.60 / $3.00 | $0.00 / $0.00 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- Kimi K2.5 public-lane score: 40.7 (Estimated · #66/128)
- Sarvam 30B public-lane score: Coming soon

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.0 | 50.8% | Coming soon | Coming soon |
| BrowseComp | 60.6% | 35.5% | Kimi K2.5 |
| OSWorld-Verified | 63.3% | Coming soon | Coming soon |
| Claw-Eval | 52.3% | Coming soon | Coming soon |
| QwenClawBench | 54.3% | Coming soon | Coming soon |
| τ³-bench results | 65.7% | Coming soon | Coming soon |
| DeepSearchQA | 77.1% | Coming soon | Coming soon |
| DeepPlanning | 14.4% | Coming soon | Coming soon |
| Toolathlon | 27.8% | Coming soon | Coming soon |
| MCP Atlas | 29.5% | Coming soon | Coming soon |
| MCP-Tasks | 59.1% | Coming soon | Coming soon |
| WideResearch | 72.7% | Coming soon | Coming soon |
| BFCL v4 | 68.3% | Coming soon | Coming soon |
| Gert Labs | 45.88% | Coming soon | Coming soon |
| ResearchClawBench | 14.0% | Coming soon | Coming soon |
| JobBench | 8.7% | Coming soon | Coming soon |

### Coding

- Winner: Not comparable
- Kimi K2.5 public-lane score: 38.5 (Estimated · #68/150)
- Sarvam 30B public-lane score: Coming soon

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 76.8% | 34% | Kimi K2.5 |
| SWE-bench Verified* | 70.8% | Coming soon | Coming soon |
| LiveCodeBench v6 | 85.0% | Coming soon | Coming soon |
| SWE-bench Pro | 50.7% | Coming soon | Coming soon |
| SWE Multilingual | 73% | Coming soon | Coming soon |
| SWE-Rebench | 58.5% | Coming soon | Coming soon |
| React Native Evals | 77.2% | Coming soon | Coming soon |
| NL2Repo | 32% | Coming soon | Coming soon |
| SciCode | 48.7% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | 50.8% | Coming soon | Coming soon |
| HumanEval | Coming soon | 92.1% | Coming soon |

### Reasoning

- Winner: Not comparable
- Kimi K2.5 public-lane score: 72.2 (Unranked · 3 rankable rows)
- Sarvam 30B public-lane score: 25.9 (Unranked · 3 rankable rows)

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| LongBench v2 | 61% | Coming soon | Coming soon |
| GPQA Diamond | Coming soon | 66.5% | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- Kimi K2.5 public-lane score: 70.4 (#32/54)
- Sarvam 30B public-lane score: Coming soon

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| MMMU | 84.3% | Coming soon | Coming soon |
| MMMU-Pro | 78.5% | Coming soon | Coming soon |
| RealWorldQA | 81.0% | Coming soon | Coming soon |
| OmniDocBench 1.5 | 88.8% | Coming soon | Coming soon |
| Video-MME | 87.4% | Coming soon | Coming soon |
| Video-MME (with subtitle) | 87.4% | Coming soon | Coming soon |
| Video-MME (w/o subtitle) | 83.2% | Coming soon | Coming soon |
| MathVision | 84.2% | Coming soon | Coming soon |
| We-Math | 84.7% | Coming soon | Coming soon |
| DynaMath | 84.4% | Coming soon | Coming soon |
| SimpleVQA | 71.2% | Coming soon | Coming soon |
| MMLongBench-Doc | 58.5% | Coming soon | Coming soon |
| CC-OCR | 79.7% | Coming soon | Coming soon |
| AI2D_TEST | 90.8% | Coming soon | Coming soon |
| CountBench | 94.1% | Coming soon | Coming soon |
| RefCOCO (avg) | 87.8% | Coming soon | Coming soon |
| MMVU | 80.4% | Coming soon | Coming soon |
| VideoMMMU | 86.6% | Coming soon | Coming soon |
| MLVU (M-Avg) | 85.0% | Coming soon | Coming soon |
| CharXiv | 77.5% | Coming soon | Coming soon |

### Knowledge

- Winner: Directional only
- Kimi K2.5 public-lane score: 51.4 (Estimated · #75/184)
- Sarvam 30B public-lane score: 27.2 (Supported · #165/184)

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 87.6% | Coming soon | Coming soon |
| GPQA-D | 87.6% | Coming soon | Coming soon |
| SuperGPQA | 69.2% | Coming soon | Coming soon |
| MMLU-Pro | 87.1% | 80% | Kimi K2.5 |
| MMLU-Pro (Arcee) | 87.1% | Coming soon | Coming soon |
| MMLU-Redux | 94.5% | Coming soon | Coming soon |
| C-Eval | 94% | Coming soon | Coming soon |
| HLE | 30.1% | Coming soon | Coming soon |
| MMLU | Coming soon | 85.1% | Coming soon |

### Multilingual

- Winner: Not comparable
- Kimi K2.5 public-lane score: 87.3 (#8/17)
- Sarvam 30B public-lane score: Coming soon

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| MMLU-ProX | 82.3% | Coming soon | Coming soon |
| NOVA-63 | 56.0% | Coming soon | Coming soon |
| INCLUDE | 83.3% | Coming soon | Coming soon |
| PolyMath | 43.1% | Coming soon | Coming soon |
| MAXIFE | 72.8% | Coming soon | Coming soon |

### Instruction Following

- Winner: Directional only
- Kimi K2.5 public-lane score: 85.8 (#48/148)
- Sarvam 30B public-lane score: 28.8 (#140/148)

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| IFEval | 93.9% | Coming soon | Coming soon |

### Mathematics

- Winner: Not comparable
- Kimi K2.5 public-lane score: 83.3 (Unranked · 7 rankable rows)
- Sarvam 30B public-lane score: Coming soon

| Benchmark | Kimi K2.5 | Sarvam 30B | Winner |
|-----------|-----------|-----------|--------|
| AIME 2025 | 96.1% | Coming soon | Coming soon |
| AIME26 | 95.8% | Coming soon | Coming soon |
| AIME25 (Arcee) | 96.3% | Coming soon | Coming soon |
| HMMT Feb 2025 | 95.4% | 73.3% | Kimi K2.5 |
| HMMT Nov 2025 | 91.1% | 74.2% | Kimi K2.5 |
| HMMT Feb 2026 | 87.1% | Coming soon | Coming soon |
| MMAnswerBench | 81.8% | Coming soon | Coming soon |
| MATH-500 | Coming soon | 97% | Coming soon |

## FAQ

### Which is better overall, Kimi K2.5 or Sarvam 30B?

Kimi K2.5 has the higher overall point estimate. Conditional score ranges do not establish rank confidence.

### Where is the biggest gap between Kimi K2.5 and Sarvam 30B?

The widest category gap is in instruction following, where the averages are 85.8 for Kimi K2.5 and 28.8 for Sarvam 30B.

### How many benchmarks does Kimi K2.5 cover on BenchLM?

Kimi K2.5 currently has 68 sourced benchmark scores on BenchLM.

### How many benchmarks does Sarvam 30B cover on BenchLM?

Sarvam 30B currently has 9 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Kimi K2.5 vs Kimi K2.5 (Reasoning)](/compare/kimi-k2-5-vs-kimi-k2-5-reasoning)
- [Sarvam 30B vs Kimi K2.5 (Reasoning)](/compare/kimi-k2-5-reasoning-vs-sarvam-30b)
- [Kimi K2.5 vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-kimi-k2-5)
- [Sarvam 30B vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-sarvam-30b)

## Explore More

- [Kimi K2.5 profile](/models/kimi-k2-5)
- [Sarvam 30B profile](/models/sarvam-30b)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)