LLM Benchmark Leaderboard
Current benchmark results across models, with notes on what each benchmark does and does not measure.
LLM Benchmark Leaderboard
Benchmark scores for 11 major language models, compiled July 2026 from public leaderboards and vendor announcements. Blank cells mean no published score — scores are never estimated. Sortable by provider and evaluation metric.
Score colours: ≥ 90 excellent 75 – 89 good < 75 below avg — not reported
11 models
| Model | Params | SWE-bench V | SWE-bench Pro | GPQA-D | OSWorld | Terminal-Bench | ARC-AGI-2 | Context | Notes |
|---|---|---|---|---|---|---|---|---|---|
Claude Fable 5 Anthropic · 2026 | Unknown | 95.0 | — | — | — | — | — | 1M | Anthropic premium tier — leads SWE-bench Verified among generally available models |
GPT-5.5 OpenAI · 2026 | Unknown | 88.7 | — | 93.6 | — | — | — | 1M | SWE-bench Verified score is vendor-reported |
Claude Opus 4.8 Anthropic · 2026 | Unknown | 88.6 | 69.2 | 93.6 | — | — | — | 1M | Leads SWE-bench Pro among active models |
Claude Sonnet 5 Anthropic · 2026-06 | Unknown | 82.1 | 63.2 | 96.2 | 88.3 | — | 84.7 | 1M | GPQA Diamond record holder (96.2%) at mid-tier pricing |
Gemini 3.1 Pro Google · 2026 | Unknown | 80.6 | — | 94.3 | — | — | 77.1 | 1M | Strong multimodal reasoning; video input |
DeepSeek V4 Flash DeepSeek · 2026-04 | 284B MoE (13B active) | 73.7 | — | 86.0 | — | — | — | 1M | Open weights (MIT); 79.0% SWE-bench Verified at max reasoning effort |
DeepSeek V4 Pro DeepSeek · 2026-04 | 1.6T MoE (49B active) | 73.6 | — | — | — | — | — | 1M | Open weights (MIT); 80.6% SWE-bench Verified at max reasoning effort |
Claude Haiku 4.5 Anthropic · 2025-10 | Unknown | 73.3 | — | — | — | — | — | 200K | Fastest and cheapest Claude |
o3 OpenAI · 2025-04 | Unknown | 69.1 | — | 83.3 | — | — | — | 200K | Scores from OpenAI's April 2025 announcement |
GPT-5.4 OpenAI · 2026 | Unknown | — | — | — | 75.0 | — | — | 1M | Value flagship with tool search |
Gemini 3.5 Flash Google · 2026 | Unknown | — | 55.1 | — | — | 76.2 | — | 1M | Agentic coding at fast-tier pricing |
Data as of May 2025. Scores may vary by source, prompt format, and evaluation methodology. Always verify against primary sources.
What the common benchmarks actually test
- MMLU / MMLU-Pro —Multiple-choice knowledge across academic subjects. Measures recall and elimination more than reasoning; largely saturated in its original form.
- GPQA —Graduate-level science questions written to be resistant to web search. Harder to contaminate and correspondingly more informative.
- SWE-bench —Resolving real GitHub issues in real repositories, verified by the project’s own tests. Among the most practically relevant benchmarks for coding agents.
- HumanEval / MBPP —Short self-contained function-writing tasks. Long saturated and a poor proxy for working in an existing codebase.
- MATH / AIME —Competition mathematics. Genuinely discriminating, and unrepresentative of most commercial work.
- LMArena —Human pairwise preference at scale. Captures things automated benchmarks miss, and rewards style and agreeableness alongside correctness.
- Needle-in-a-haystack —Retrieving a planted fact from a long context. Necessary but easy — real long-context work requires synthesis across many passages, which these tests do not measure.
About
This leaderboard aggregates published benchmark scores for leading large language models across six current evaluations: SWE-bench Verified (real-world software engineering), SWE-bench Pro (contamination-resistant coding), GPQA Diamond (PhD-level science), OSWorld-Verified (computer use), Terminal-Bench 2.1 (agentic terminal work), and ARC-AGI-2 (abstract reasoning). Scores are sourced from public leaderboards and vendor announcements as of July 2026 — blank cells mean no published score, never an estimate. Filter by provider to narrow the comparison, click column headers to sort.
How to use
- 1 Use the provider filter chips to show only models from specific vendors.
- 2 Click any column header to sort by that benchmark — click again to reverse.
- 3 Score cells are color-coded: green (≥90), amber (75–89), muted (<75).
- 4 Missing scores are shown as "—" (not all models are evaluated on all benchmarks).
- 5 Click "What is X?" links to expand descriptions of each benchmark.
- What is SWE-bench Verified?
- SWE-bench Verified is a set of 500 human-validated software engineering tasks taken from real GitHub issues in popular Python repositories. The model must produce a code patch that passes the repository's own test suite. It has become the standard measure of practical, agentic coding ability — as of July 2026 the top published score is 95% (Claude Fable 5).
- What is GPQA Diamond?
- GPQA Diamond is a set of 198 graduate-level physics, chemistry, and biology questions written by domain experts and designed to be "Google-proof" — skilled humans with full web access score only ~34%. Frontier models now exceed 90%, making it one of the few knowledge benchmarks that still differentiates top models.
- Are these benchmarks still reliable?
- Benchmark saturation is why this leaderboard tracks the current generation of evaluations: classics like MMLU, HumanEval, and GSM8K are effectively solved (top models cluster above 90%), so the industry moved to harder, contamination-resistant tests like SWE-bench Pro and ARC-AGI-2. Even so, vendor-reported scores can use favourable scaffolding — for robust comparisons, combine benchmarks with task-specific evaluations on your own data.
More in AI Tools
Token Counter Count tokens for any text using the real tokenizers, with a per-token breakdown showing exactly where a passage splits.LLM Cost Calculator Estimate API spend from input and output token volumes across providers, and compare the same workload model by model.Context Window Planner Check whether a system prompt, retrieved documents, history and expected output all fit inside a model’s window.Model Comparison Compare models side by side on context length, pricing, modality and benchmark results.Prompt Builder Compose structured prompts from role, context, task and format sections, with a live token count as you write.Prompt Diff Compare two prompt versions token by token to see exactly what changed between runs.
See all ai tools.