Grok 4.7 vs Kimi K3.1 Pro
Full profiles: Grok 4.7 (SpaceXAI) and Kimi K3.1 Pro (Moonshot AI): every benchmark, price and rank.
Updated · Data from 6 sources
As of 24 Oct 2026, Grok 4.7 (high) and Kimi K3.1 Pro are neck and neck on quality (63.9 vs 66.8 on the BenchLeader Index, inside the ±9.6 margin of error), and Grok 4.7 (high) costs 1.3× less, which makes it the better buy for most workloads. Kimi K3.1 Pro leads in every category we measure both in.
Grok 4.7 benchmarks 
Kimi K3.1 Pro benchmarks Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Reasoning | Kimi K3.1 Pro | 81 vs 61 category score; CritPt 24.6% vs 18.0% | High |
| Knowledge | Either | 75 vs 75 category score; AA-Omniscience 30.9 vs 29.5 | Either |
| Long documents | Kimi K3.1 Pro | 68 vs 64 category score; AA-LCR 86.3% vs 77.0% | Medium |
| Lowest cost | Grok 4.7 | $3.00 vs $4.00 per 1M blended; agentic coding session $0.144 vs $0.160 | – |
| Fastest first answer | Grok 4.7 | 19.6 s (Grok 4.7 (high)) vs 27.7 s (Kimi K3.1 Pro) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Grok 4.7 (high) and Kimi K3.1 Pro are neck and neck on quality (63.9 vs 66.8, within the ±9.6 margin).
- Grok 4.7 (high) is stronger in coding, composite, knowledge.
- Kimi K3.1 Pro is stronger in long context, reasoning, agents & tools, multimodal.
- Grok 4.7 (high) is 1.3× cheaper ($3.00 vs $4.00 per 1M blended).
At a glance
| Spec | Grok 4.7 | Kimi K3.1 Pro |
|---|---|---|
| Developer | SpaceXAI | Moonshot AI |
| Released | 21 Sept 2026 | 23 Oct 2026 |
| Context window (tokens) | 500,000 | 1,048,576 |
| Compared at | Grok 4.7 (high) | Kimi K3.1 Pro |
| Open weights | No | No |
| Price in / out / cached input per 1M | $2.00 / $6.00 / $0.500 | $2.00 / $10.00 / $0.200 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Grok 4.7 (high) | Kimi K3.1 Pro | Leader |
|---|---|---|---|
| Reasoning | 60.9 | 81.3 | Kimi K3.1 Pro |
| Coding | 64.8 | – | not comparable |
| Agents & tools | – | 67.3 | not comparable |
| Knowledge | 75.4 | 74.8 | level |
| Multimodal | – | 67.0 | not comparable |
| Long context | 63.5 | 68.3 | Kimi K3.1 Pro |
| Composite | 81.3 | 75.8 | Grok 4.7 |
Where sources disagree
Across the 12 benchmarks both have, Grok 4.7 leads 2, Kimi K3.1 Pro leads 10. Every overall index that rates both puts Kimi K3.1 Pro ahead.
| Overall index | Grok 4.7 (high) | Kimi K3.1 Pro | Leader |
|---|---|---|---|
| BenchLeader Index | 63.9 | 66.8 | neck and neck |
| AA Intelligence Index v4.3.2 | 46.3 | 55.1 | Kimi K3.1 Pro |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.
| Measure | Grok 4.7 (high) | Kimi K3.1 Pro |
|---|---|---|
| Blended price per 1M tokens | $3.00 | $4.00 |
| Input / output per 1M | $2.00 / $6.00 | $2.00 / $10.00 |
| Cached input per 1M | $0.500 | $0.200 |
| Price per task: chat reply | $0.0026 | $0.0038 |
| Price per task: summarize a 30-page report | $0.028 | $0.030 |
| Price per task: code edit | $0.021 | $0.027 |
| Price per task: agentic coding session | $0.144 | $0.160 |
| Price per task: structured extraction | $0.0052 | $0.0060 |
| Cost to run a full benchmark suite | $2.73 | $2.11 |
| Output speed | 91 tok/s | 81 tok/s |
| Time to first answer | 19.6 s | 27.7 s |
| Time to a complete answer | 25.1 s | 33.9 s |
Per token is not per task: on Artificial Analysis' full suite Grok 4.7 (high) cost $2.73 and Kimi K3.1 Pro $2.11, because Grok 4.7 (high) used about 1.7× as many tokens for the same work, which a per-token price hides.
Each model is compared at its best-ranked reasoning-effort setting. Grok 4.7 (high) scores 63.9 and takes 20 s to its first answer token; Grok 4.7 (xhigh) scores 61.6 in 35 s, Grok 4.7 (low) scores 61.0 in 4 s. Kimi K3.1 Pro scores 66.8 and takes 28 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
109 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Grok 4.7 (high) | Kimi K3.1 Pro |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 63.9 | 66.8 |
| Coding score | BenchLeader | 64.8 | – |
| Composite score | BenchLeader | 81.3 | 75.8 |
| Knowledge score | BenchLeader | 75.4 | 74.8 |
| Long context score | BenchLeader | 63.5 | 68.3 |
| Reasoning score | BenchLeader | 60.9 | 81.3 |
| Agents & tools score | BenchLeader | – | 67.3 |
| Multimodal score | BenchLeader | – | 67.0 |
| Blended price $/M | BenchLeader | $3.00 | $4.00 |
| Output speed | BenchLeader | 91 tok/s | 81 tok/s |
| Time to first answer | BenchLeader | 19.6 s | 27.7 s |
| Context window | BenchLeader | 500,000 tokens | 1,048,576 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 92.7%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1464† | – |
| LiveBench 2.0 Reasoning | LiveBench | 82.7%† | 91.1% |
| LiveBench 1.0 Reasoning | LiveBench | 82.7%† | – |
| Humanity's Last Exam (AA) | Artificial Analysis | 42.3% | 52.5% |
| ARC-AGI-1 | ARC Prize | 90.2% | – |
| ARC-AGI-2 | ARC Prize | 58.3% | – |
| ARC-AGI-3 | ARC Prize | 10.0% | – |
| CritPt | Artificial Analysis | 18.0% | 24.6% |
| MysteryMechanism | Vals AI | 25.2%† | – |
| Chess Puzzles | Epoch AI Benchmarking Hub | 41.0%† | – |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 29.0%† | – |
| LMCA | Epoch AI Benchmarking Hub | 49.4%† | – |
| DTBench | Epoch AI Benchmarking Hub | 96.0%† | – |
| Coding | |||
| SciCode | SciCode | 57.8% | – |
| FrontierCode | Cognition | 47.6%† | – |
| Arena Coding | Arena (LMArena) | 1493† | – |
| Arena WebDev | Arena (LMArena) | 1637† | – |
| LiveBench 2.0 Coding | LiveBench | 84.5%† | 81.8% |
| LiveBench 1.0 Coding | LiveBench | 77.2%† | – |
| SciCode (AA) | Artificial Analysis | 57.8% | 60.8% |
| IOI | Vals AI | 57.7%† | – |
| Code Migration | Vals AI | 44.8%† | – |
| Code Migration v1.1 | Vals AI | 53.2%† | – |
| ProgramBench | Vals AI | 0.5%† | – |
Show all 109 benchmarks
| Metric | Publisher | Grok 4.7 (high) | Kimi K3.1 Pro |
|---|---|---|---|
| Coding | |||
| SWE-bench Multimodal | Vals AI | 42.9%† | – |
| Vibe Code Bench v1.1 | Vals AI | 86.2%† | – |
| CursorBench | Cursor | 43.9% | – |
| FrontierSWE | Epoch AI Benchmarking Hub | 29.5%† | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 24.8% | 67.2% |
| Startrise Frontend | Startrise | 75.9%† | – |
| FrontierCode 1.1 (AA) | Artificial Analysis | 34.1% | 47.1% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 37.6%† | – |
| APEX-Agents | Mercor | 54.6%† | – |
| Arena Agent | Arena (LMArena) | 3.2† | – |
| LiveBench Long Agentic | LiveBench | 50.1%† | 49.5% |
| LiveBench Browser | LiveBench | 60.0%† | 80.0% |
| LiveBench Agentic Coding | LiveBench | 54.0%† | – |
| Terminal-Bench 2.1 (Vals) | Vals AI | 73.4%† | – |
| GDPval-AA v2.1 | Artificial Analysis | 60.4% | 67.3% |
| ITBench SRE (AA) | Artificial Analysis | 42.1%† | – |
| Analyst Agent (AA) | Artificial Analysis | 55.0% | – |
| CyberBench | Vals AI | 69.5%† | – |
| Finance Agent v2 | Vals AI | 52.3%† | – |
| Finance Agent v2.1 | Vals AI | 60.6%† | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 18.3%† | – |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 12.5%† | – |
| Legal Research Bench | Vals AI | 47.1%† | – |
| Robo-bench | Vals AI | 0.2%† | – |
| SREBench | Vals AI | 3.8%† | – |
| Tax Agent Bench | Vals AI | 34.2%† | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 28.8%† | – |
| Terminal-Bench Science | Vals AI | 10.0%† | – |
| Vending-Bench 2 | Andon Labs | 10536.8† | – |
| AutomationBench | Artificial Analysis | 63.5% | 73.3% |
| GDP.pdf (AA) | Artificial Analysis | 23.2% | 31.4% |
| Harvey LAB-AA v1.1 | Harvey | 9.4%† | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1639† | 1678 |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 9.5% | – |
| CWE-Bench (AA) | Artificial Analysis | 68.3%† | 60.8% |
| DeepSec-Bench (AA) | Artificial Analysis | 26.9%† | 36.7% |
| CyberGym end-to-end (AA) | Artificial Analysis | 74.0%† | 94.7% |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 53.0%† | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 17.1%† | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 98.1%† | – |
| ProofBench | Vals AI | 26.0%† | – |
| LiveBench Mathematics | LiveBench | 95.7%† | – |
| Arena Maths | Arena (LMArena) | 1443† | – |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 56.0%† | – |
| LiveBench 2.0 Data Analysis | LiveBench | 76.9%† | 79.1% |
| LiveBench Research | LiveBench | 86.4%† | 77.9% |
| LiveBench 1.0 Data Analysis | LiveBench | 76.9%† | – |
| AA-Omniscience | Artificial Analysis | 30.9 | 29.5 |
| LegalBench | Vals AI | 84.4%† | – |
| PRBench Legal | Scale AI SEAL | 47.6%† | – |
| BioMysteryBench | Vals AI | 69.3%† | – |
| Excel Modeling Benchmark | Vals AI | 67.0%† | – |
| Excel Modeling Benchmark v1.1 | Vals AI | 71.5%† | – |
| MedCode | Vals AI | 49.5%† | – |
| MedScribe | Vals AI | 89.4%† | – |
| Public Benefits Bench v1.1 | Vals AI | 65.6%† | – |
| GDP.pdf | Epoch AI Benchmarking Hub | 22.8%† | – |
| AA-Omniscience: accuracy | Artificial Analysis | 47.8% | 51.0% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 67.6% | 56.1% |
| Instruction following | |||
| LiveBench Language | LiveBench | 80.1%† | – |
| LiveBench Instruction Following | LiveBench | 75.3%† | – |
| Arena Instruction Following | Arena (LMArena) | 1440† | – |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1441† | – |
| Arena Creative Writing | Arena (LMArena) | 1418† | – |
| Arena Multi-turn | Arena (LMArena) | 1431† | – |
| Arena Longer Queries | Arena (LMArena) | 1453† | – |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1231† | – |
| LiveBench Design | LiveBench | 78.0%† | 77.8% |
| MMMU-Pro | Artificial Analysis | 78.2%† | 83.9% |
| SAGE | Vals AI | 40.8%† | – |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 32.5%† | – |
| AA-3D | Artificial Analysis | 995† | – |
| Long context | |||
| AA-LCR | Artificial Analysis | 77.0% | 86.3% |
| MLCR | Artificial Analysis | 18.9% | – |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 154.4† | – |
| LiveBench 2.0 | LiveBench | 74.1%† | 76.7% |
| LiveBench 1.0 | LiveBench | 77.4%† | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 46.3 | 55.1 |
| Vals Index | Vals AI | 58.6† | – |
| Vals Finance Index | Vals AI | 58.5%† | – |
| Vals RSI Index | Vals AI | 20.2%† | – |
| Vals Science Index | Vals AI | 34.8%† | – |
| Vals Router Index | Vals AI | 58.6%† | – |
| AA Cyber Index | Artificial Analysis | 56.4† | 64.1 |
Data as of 2026-10-24. Best configuration of each model; every score links to its source on the model pages.
What changed
- Kimi K3.1 Pro: BenchLeader Index 65.6 → 66.8.
- Kimi K3.1 Pro: 5 new results (FrontierCode 1.1 (AA), AA Cyber Index, CWE-Bench (AA)…).
- Kimi K3.1 Pro: first results.
- Grok 4.7 (high): 1 new result (MLCR).
- Grok 4.7 (high): 4 new results (AutomationBench, GDP.pdf (AA), FrontierCode 1.1 (AA)…).
- Grok 4.7 (high): 1 new result (Analyst Agent (AA)).
- Grok 4.7 (high): BenchLeader Index 63.6 → 64.3.
- Grok 4.7 (high): 3 new results (ARC-AGI-1, ARC-AGI-2, ARC-AGI-3).
- Grok 4.7 (high): BenchLeader Index 65.8 → 63.6.
- Grok 4.7 (high): 1 new result (SciCode).
Related comparisons
- Grok 4.7 vs Claude Opus 5.5
- Grok 4.7 vs Gemini 4 Argon
- Grok 4.7 vs Muse Wonder Full
- Kimi K3.1 Pro vs Claude Opus 5.5
- Kimi K3.1 Pro vs Gemini 4 Argon
- Kimi K3.1 Pro vs Muse Wonder Full
- Kimi K3.1 Pro vs Claude Fable 5.1
- Grok 4.6 vs Kimi K3.1 Pro
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Grok 4.7 vs Kimi K3.1 Pro: questions
- Is Grok 4.7 better than Kimi K3.1 Pro?
- Grok 4.7 (high) and Kimi K3.1 Pro are neck and neck on quality (63.9 vs 66.8, within the ±9.6 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-24, so check the category scores for your use.
- Which is cheaper, Grok 4.7 or Kimi K3.1 Pro, and will that change?
- Grok 4.7 is cheaper today: $3.00 against $4.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Which is better for long documents?
- Kimi K3.1 Pro scores higher on long-context benchmarks (68 vs 64). Kimi K3.1 Pro accepts more context: 1,048,576 against 500,000 tokens for Grok 4.7.
- Which has lower latency?
- Grok 4.7 (high) answers first: 19.6 s against 27.7 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Which is faster, Grok 4.7 or Kimi K3.1 Pro?
- Grok 4.7 streams faster: 91 against 81 output tokens per second.
- Do independent benchmarks agree on Grok 4.7 vs Kimi K3.1 Pro?
- Across the 12 benchmarks both have, Grok 4.7 leads 2, Kimi K3.1 Pro leads 10. Every overall index that rates both puts Kimi K3.1 Pro ahead.
- When was Kimi K3.1 Pro released?
- Kimi K3.1 Pro was released on 23 Oct 2026.