Z.aiLanguage model
GLM-4.7-FlashX
GLM-4.7-FlashX has no overall index value yet because there are too few independent benchmark results.
28out of 100same as last month
based on two data groups, mostly search
Summary
GLM-4.7-FlashX is a separate API entry in the GLM-4.7 series with text as both input and output and a 200K context. Z.ai publishes no separate FlashX weight card. An identity with GLM-4.7-Flash is therefore unverified.
FlashX therefore remains an API entry without a documented local checkpoint or its own parameter size. The open Flash variant cannot be treated as its weight identity.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 12 published benchmark results
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
LawBench (Mem) (Show measurement, test conditions, and source)
- Source value
- 41.77
- Score
- 41.77
- Metric
- Mem (Memorization)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LawBench. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LawBench (Und) (Show measurement, test conditions, and source)
- Source value
- 41.9
- Score
- 41.9
- Metric
- Und (Understanding)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LawBench. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LawBench (App) (Show measurement, test conditions, and source)
- Source value
- 56.28
- Score
- 56.28
- Metric
- App (Application)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LawBench. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LawBench (Avg) (Show measurement, test conditions, and source)
- Source value
- 47.64
- Score
- 47.64
- Metric
- Avg (Average score)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LawBench. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Mem) (Show measurement, test conditions, and source)
- Source value
- 42.49
- Score
- 42.49
- Metric
- Mem (Memorization)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Und) (Show measurement, test conditions, and source)
- Source value
- 81.89
- Score
- 81.89
- Metric
- Und (Understanding)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Log) (Show measurement, test conditions, and source)
- Source value
- 54.41
- Score
- 54.41
- Metric
- Log (Logical inference)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Dis) (Show measurement, test conditions, and source)
- Source value
- 27.3
- Score
- 27.3
- Metric
- Dis (Discrimination)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Gen) (Show measurement, test conditions, and source)
- Source value
- 23.09
- Score
- 23.09
- Metric
- Gen (Generation)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Eth) (Show measurement, test conditions, and source)
- Source value
- 43.13
- Score
- 43.13
- Metric
- Eth (Ethics)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
LexEval (Avg) (Show measurement, test conditions, and source)
- Source value
- 49.55
- Score
- 49.55
- Metric
- Avg (Average score)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on LexEval. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
PoliLegal (Show measurement, test conditions, and source)
- Source value
- 72.5
- Score
- 72.5
- Metric
- Accuracy (unified metric for the PoliLegal benchmark, per section 5.1.2)
- Unit
- No unit provided
- Direction
- No direction published
- Source type
- independent-evaluator
- Evaluator
- PoliLegalLM report, arXiv:2604.17543v1
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Independent evaluation. Table 3 row `GLM-4.7-FlashX`; Params `30B/A3B`; Local `Yes`. Section 5.1.3 also lists `GLM-4.7-FlashX` among comparison baselines. The report compares the model on PoliLegal. The source does not state a numeric unit, evaluated N, confidence interval, or sampling/decoding settings for this result.
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
GLM-4.7-FlashXTrinity Large Thinking 28MiMo-V2.5-Pro-UltraSpeed 28Nex-N2.5-Max 28OLMo 2 13B 28
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as GLM-4.7-FlashX, chosen by model line, developer, size, price and release date.
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Prices
Not included in any subscription in Germany.
Frequently asked questions about GLM-4.7-FlashX
What is the knowledge cutoff of GLM-4.7-FlashX?
April 2025, according to Z.ai. That is 9 months before release.
Stated by Z.ai
How large is the context window of GLM-4.7-FlashX?
200,000 tokens. That is how much text GLM-4.7-FlashX reads at once, about 300 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Can I run GLM-4.7-FlashX locally?
Yes, the weights are open.
Stated by Z.ai
- Open weights
- yes
- License
- MIT
- Parameters
- 30B
- active per token
- 3B
Sources and data date
Show all 14 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Input | Documented modalities. Retrieved October 27, 2026. |
| Output | Documented modalities. Retrieved October 27, 2026. |
| Research status | Research date October 15, 2026. 6 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Unresolved | exact-flashx-checkpoint-scores Note The official card provides GLM-4.7-Flash scores, but neither this card nor the reviewed source establishes that the catalog's FlashX API variant is the same checkpoint. No sibling score is transferred. |
| Not found in the inspected sources | exact-flashx-quality-benchmarks Note No exact GLM-4.7-FlashX numeric quality result with a sourceable benchmark protocol was found. Public material for the related GLM-4.7-Flash model is not transferred, and third-party comparisons without identifiable benchmark provenance are not treated as scores. |
| Unresolved | quality-benchmarks Note No numeric evaluation with a reproducible configuration was found for the exact GLM-4.7-FlashX route. GLM-4.7-Flash is a different variant and its values are not transferred. |
| Unresolved | evaluation configuration and uncertainty Original wording The captured paper table supplies the exact model identity, Params 30B/A3B, Local Yes, scores and task labels, but does not report result units, per-result evaluated N, confidence intervals, or sampling/decoding settings. Dataset sizes mentioned elsewhere in the paper are not used as evaluated N. |
| Additional source | Hugging Face (retrieved October 15, 2026) · GLM-4.7-Flash model card · Editorial description reviewed October 2, 2026 |
| Additional source | PoliLegalLM report, arXiv:2604.17543v1 (retrieved October 18, 2026) |
| Additional source | Model metadata (retrieved October 16, 2026)123 |
| Additional source | Model metadata (retrieved October 16, 2026) · Z.AI GLM-4.7 documentation · Editorial description reviewed October 2, 2026 |
| Additional source | GLM-4.7-Flash model card text · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |