SpaceXAIGrok 4.5
Grok 4.5 is a SpaceXAI proprietary reasoning model, released 8 Jul 2026. Its best configuration (high reasoning effort) ranks #134 of 785 on the BenchLeader Index at 59.3 ±4.6, in the upper half. The ± is the point: 233 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in composite (73) and lowest in agents & tools (50). At $3.00 per million tokens blended it is pricier than most ranked models. Output speed of 72 tokens per second puts it slower than most, with a first answer in 9.2 s. It has been measured at 4 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 20 Oct 2026.
- Blended price
- $3.00/M
- $2.00 in · $6.00 out
- Output speed
- 72 tok/s
- measured by Artificial Analysis
- First answer
- 9.17 s
- first token 1.66 s
- Context
- 500k
- Full answer
- 16 s
- median, reasoning included
- Cost per run
- $0.881
- one full Intelligence Index run
- Released
- 8 Jul 2026
- Overall index59
- Reasoning60
- Coding57
- Agents & tools50
- Maths53
- Knowledge63
- Multimodal64
- Long context65
- Composite73
Versions
SpaceXAI has shipped 11 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Grok 4.7high | 21 Sept 2026 | 63.9 | #59 |
| Grok 4.6medium | 11 Aug 2026 | 64.2 | #56 |
| Grok 4.5highthis page | 8 Jul 2026 | 59.3 | #134 |
| Grok 4.3medium | 17 Apr 2026 | 58.7 | #147 |
| Grok 4.2high | 10 Mar 2026 | – | – |
| Grok 4.20thinking | 17 Feb 2026 | 58.6 | #148 |
| Grok 4.1 | 17 Nov 2025 | 57.2 | #184 |
| Grok 4 | 9 Jul 2025 | 58.0 | #161 |
| Grok 3 | 17 Feb 2025 | 51.5 | #340 |
| Grok 2 | 12 Dec 2024 | 41.9 | #612 |
| Grok 4.20 1 | – | – | – |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| low | – | – | 55 tok/s | 1.66 s | $0.0026 | Reasoning 48 |
| medium | – | – | 55 tok/s | 1.66 s | $0.0026 | Reasoning 54 |
| highbest | 59.3 | #134 | 72 tok/s | 9.17 s | $0.0026 | Agents & tools 50 · Coding 57 · Composite 73 · Knowledge 63 · Long context 65 · Maths 53 · Multimodal 64 · Reasoning 60 |
| not stated | 59.0 | #140 | 55 tok/s | 1.66 s | $0.0026 | Agents & tools 57 · Coding 57 · Composite 53 · Human preference 65 · Maths 65 · Multimodal 63 · Reasoning 69 |
How its index has moved
8 Sept 2026 to 23 Oct 2026- high
- not stated
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | low | medium | high | not stated | Source | Trend |
|---|---|---|---|---|---|---|
| GPQA Diamond | – | – | 93.4%#17 | – | Epoch AI Benchmarking Hub | |
| SimpleBencheffort not stated | – | – | – | 70.0%#14 | SimpleBench | |
| Arena Hard Promptseffort not stated | – | – | – | 1488#52 | Arena (LMArena) | |
| LiveBench 2.0 Reasoningnot in indexeffort not stated | – | – | – | 87.2%#26 | LiveBench | |
| LiveBench 1.0 Reasoningnot in indexeffort not stated | – | – | – | 87.2%#29 | LiveBench | |
| GPQA Diamond (AA)not in index | – | – | 93.1%#24 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | – | 42.7%#65 | – | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | – | 92.9%#12 | – | Vals AI | |
| Kagi LLM Benchmarkeffort not stated | – | – | – | 83.5%#6 | Kagi LLM Benchmark | |
| ARC-AGI-1 | 79.2%#122 | 87.2%#97 | 85.7%#110 | – | ARC Prize | |
| ARC-AGI-2 | 33.1%#126 | 52.6%#106 | 52.6%#106 | – | ARC Prize | |
| ARC-AGI-3 | 0.3%#53 | 0.3%#53 | 0.3%#53 | – | ARC Prize | |
| CritPt | – | – | 15.4%#89 | – | Artificial Analysis | |
| Chess Puzzlesnot in index | – | – | 35.0%#57 | – | Epoch AI Benchmarking Hub | |
| LMCAnot in index | – | – | 45.2%#89 | – | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | – | – | 96.5%#22 | – | Epoch AI Benchmarking Hub | |
| ForecastBenchnot in index | – | – | 66.1%#6 | – | ForecastBench |
Coding
Agents & tools
Maths
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| FrontierMath Tiers 1–3 | 57.2%#45 | – | Epoch AI Benchmarking Hub | |
| FrontierMath Tier 4 | 24.4%#50 | – | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 97.8%#31 | – | Epoch AI Benchmarking Hub | |
| ProofBench | 31.0%#42 | – | Vals AI | |
| LiveBench Mathematicsnot in indexeffort not stated | – | 90.8%#34 | LiveBench | |
| Arena Mathseffort not stated | – | 1473#47 | Arena (LMArena) |
Knowledge
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| SimpleQA Verified | 48.3%#32 | – | Epoch AI Benchmarking Hub | |
| LiveBench 2.0 Data Analysisnot in indexeffort not stated | – | 73.0%#32 | LiveBench | |
| LiveBench Researchnot in indexeffort not stated | – | 85.9%#6 | LiveBench | |
| LiveBench 1.0 Data Analysisnot in indexeffort not stated | – | 73.0%#43 | LiveBench | |
| AA-Omniscience | 25.3#52 | – | Artificial Analysis | |
| MMLU-Pro | 89.2%#14 | – | Vals AI | |
| LegalBench | 86.0%#19 | – | Vals AI | |
| CorpFin | 67.4%#12 | – | Vals AI | |
| TaxEval | 71.7%#72 | – | Vals AI | |
| Excel Modeling Benchmarknot in index | 52.9%#51 | – | Vals AI | |
| MedCodenot in index | 43.3%#46 | – | Vals AI | |
| MedScribenot in index | 86.9%#19 | – | Vals AI | |
| GDP.pdfnot in index | 14.0%#42 | – | Epoch AI Benchmarking Hub | |
| AA-Omniscience: accuracynot in index | 51.5%#55 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 45.9%#141 | – | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| LiveBench Languagenot in index | 82.8%#23 | LiveBench | |
| LiveBench Instruction Followingnot in index | 71.5%#23 | LiveBench | |
| Arena Instruction Followingnot in index | 1461#50 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1466#52 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1452#36 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1473#55 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1480#49 | Arena (LMArena) |
Multimodal
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| Arena Visioneffort not stated | – | 1287#31 | Arena (LMArena) | |
| LiveBench Designnot in indexeffort not stated | – | 78.1%#19 | LiveBench | |
| MMMU-Pro | 80.4%#49 | – | Artificial Analysis | |
| MMMU-Pro (Vals)not in index | 61.8%#80 | – | Vals AI | |
| MortgageTaxnot in index | 61.8%#57 | – | Vals AI | |
| SAGEnot in index | 35.0%#69 | – | Vals AI | |
| Blueprint-Bench 2not in indexeffort not stated | – | 27.3%#20 | Epoch AI Benchmarking Hub |
Long context
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| AA-LCR | 79.3%#98 | – | Artificial Analysis | |
| Arena Documentnot in indexeffort not stated | – | 1464#22 | Arena (LMArena) | |
| MLCRnot in index | 16.1%#69 | – | Artificial Analysis |
Composite
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | 154.1#42 | Epoch AI Benchmarking Hub | |
| LiveBench 2.0effort not stated | – | 70.1%#25 | LiveBench | |
| LiveBench 1.0not in indexeffort not stated | – | 75.8%#35 | LiveBench | |
| AA Intelligence Index v4.3.2 | 38.8#75 | – | Artificial Analysis | |
| Vals Multimodal Indexnot in index | 63.4%#15 | – | Vals AI |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| SpaceXAI | 54 tok/s | 1.74 s | $2.00 | $6.00 | 500k | – |
| SpaceXAI | 51 tok/s | 1.55 s | $2.00 | $6.00 | 500k | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.300 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0026 | $0.0021 | 13.3 s |
| Summarize a 30-page report | 12,000 / 600 | $0.028 | $0.012 | 17.5 s |
| Code edit | 6,000 / 1,500 | $0.021 | $0.013 | 30.1 s |
| Agentic coding session | 60,000 / 4,000 | $0.144 | $0.068 | 1.1 min |
| Structured extraction | 2,000 / 200 | $0.0052 | $0.0027 | 12.0 s |
More from SpaceXAI
See also
Questions people ask
- What are Grok 4.5's benchmark scores?
- As of 24 Oct 2026, Grok 4.5 (high effort) scores 59.3 ±4.6 on the BenchLeader Index, strongest in composite (73) and weakest in agents & tools (50). That rests on 24 independent benchmark results, and every score on this page links to its publisher.
- How does Grok 4.5 rank?
- Grok 4.5 (high effort) is #134 of 785 ranked configurations on 24 Oct 2026. 233 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Grok 4.5 cost per 1M tokens?
- As of 24 Oct 2026, Grok 4.5 costs $2.00 per 1M input tokens and $6.00 per 1M output tokens, $3.00 blended (three input tokens per output token).
- When was Grok 4.5 released?
- SpaceXAI released Grok 4.5 on 8 Jul 2026. BenchLeader has results for it at 4 reasoning-effort settings.
Data as of 24 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Grok 4.5: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/grok-4-5.