Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Reflection AI · Open Weight · bench-align-v5
Beam 1.0 501B-A23B NVFP4 is built for reasoning. It ranks in the top quarter for math (top 17% of 111 scored models) and reasoning (top 19% of 216 scored models), by BenchLM's published scores.
Beam 1.0 501B-A23B NVFP4 by Reflection AI carries a context window of 1M tokens — observed 2026-11-21.
Every published price component, including cache reads and writes.
10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. This uses the same calculator as the cost simulator, so an unavailable applicable rate makes the example unavailable too.
Lifecycle events the source associates with this model, plus what this profile does not claim.
That is not evidence the model has no lifecycle plan — only that this source published none.
Seven axes from the ranking source. A missing axis is drawn as a gap.
Each axis shows the ranking source’s score and, beside it, BenchLM’s separate verified score for that axis. The two are published side by side and are not interchangeable.
Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim.
Time to first token: latency to the first answer chunk (Artificial Analysis, via BenchLM). Reasoning models include thinking time, so values can run to tens or hundreds of seconds.
Evidence key: ObservedUnavailable
| Measurement | Value | Observed | Last good | Evidence |
|---|---|---|---|---|
| Time to first token | Unavailable | Unavailable | Unavailable | |
| Throughput | Unavailable | Unavailable | Unavailable |
Regional or per-endpoint measurements appear only when the API supplies them; none are modelled here.
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| GPQA-D GPQA Diamond | 89.3 | #41 | Graduate-level science questions | Graduate level | GPQA: A Graduate-Level Google-Proof Q&A Benchmark |
| HLE w/o tools Humanity's Last Exam without tools | 34.8 | #36 | Expert-level questions | Frontier expert level | Introducing GPT-5.4 mini and nano |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| Terminal-Bench 2.1 Terminal-Bench 2.1 | 78.4 | #40 | Terminal-based software-agent tasks | Professional software engineering | Terminal-Bench |
| SWE-bench Verified Software Engineering Benchmark Verified | 80.7 | #16 | 500 verified issues | Professional software engineering | SWE-bench: Can Language Models Resolve Real-World GitHub Issues? |
| SWE-bench Pro SWE-bench Pro | 65.6 | #15 | 1,865 repository problems | Long-horizon professional engineering | SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks? |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| AIME26 AIME 2026 | 97.5 | #4 | Competition math problems | Olympiad-style mathematics | Qwen3.6 launch benchmarks |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| LongBench v2 LongBench v2 | 67.2 | #4 | Long-context tasks | Hard long-context | LongBench v2 |
| AA-LCR Artificial Analysis Long Context Reasoning | 78.7 | #64 | Long-context reasoning tasks | Long-context reasoning | Artificial Analysis model benchmarks |
| CritPt Critical Physics Tasks | 15.4 |
| Benchmark | Value | Rank | Tasks | Difficulty | Provenance |
|---|---|---|---|---|---|
| Terminal-Bench 2.1 Terminal-Bench 2.1 | 78.4 | #42 | Terminal-based software-agent tasks | Professional software engineering | Terminal-Bench |
| BrowseComp BrowseComp | 75.6 | #41 | Research questions requiring browsing | Hard web research | BrowseComp |
| HLE w/ tools Humanity's Last Exam with tools | 51.1 |
| SWE Multilingual SWE Multilingual | 77.5 | #22 | Multilingual software-engineering tasks | Professional software engineering | MiniMax M2.7: Early Echoes of Self-Evolution |
| SWE-Atlas Codebase QnA SWE-Atlas Codebase QnA | 31.4 | #8 | 124 codebase questions across 11 repositories | Production codebase comprehension | Muse Spark 1.3 Evaluation Methodology |
| SciCode Scientific Code Benchmark | 46.9 | #14 | — | — | BenchLM |
| DeepSWE DeepSWE | 42.7 | #62 | 113 software engineering tasks across 91 repositories and 5 languages | Long-horizon software engineering | DeepSWE benchmark blog |
| Research-level physics questions |
| Research-level physics reasoning |
| CritPt Benchmark Leaderboard |
| #27 |
| Expert questions with tool use |
| Frontier tool-augmented reasoning |
| DeepSeek-V4 Technical Report |
| MCP Atlas MCP Atlas | 80.5 | #16 | Tool-integrated agent tasks | Advanced tool use | Introducing GPT-5.4 mini and nano |
| DeepSearchQA DeepSearchQA | 81.1 | #14 | Agentic browsing and list-answer questions | Agentic web research | Muse Spark Eval Methodology |
| τ³-bench results τ³-Bench Tool-Agent-User Evaluation | 35.4 | #22 | Corrected customer-service tasks plus knowledge and voice evaluation modes | Long-horizon, multimodal, and knowledge-aware tool use | Official τ³-bench repository and release notes |