Agnes 2.5 Pro Theta vs GLM 4.7 Flash
Full profiles: Agnes 2.5 Pro Theta (Sapiens AI) and GLM 4.7 Flash (Zhipu AI): every benchmark, price and rank.
Updated · Data from 6 sources
As of 9 Nov 2026, Agnes 2.5 Pro Theta and GLM 4.7 Flash are neck and neck on quality (50.8 vs 51.1 on the BenchLeader Index, inside the ±8.7 margin), and GLM 4.7 Flash is 3.7× cheaper at $0.153 vs $0.563 per 1M tokens blended. GLM 4.7 Flash leads in every category we measure both in. Agnes 2.5 Pro Theta streams 3×+ faster.
Key facts (as of 9 Nov 2026)
- Agnes 2.5 Pro Theta scores 50.8 on the BenchLeader Index and GLM 4.7 Flash scores 51.1 (BenchLeader, 9 Nov 2026).
- Agnes 2.5 Pro Theta costs $0.563 per 1M tokens blended (list price, 9 Nov 2026).
- GLM 4.7 Flash costs $0.153 per 1M tokens blended (list price, 9 Nov 2026).
- Agnes 2.5 Pro Theta streams 3×+ faster than GLM 4.7 Flash (Artificial Analysis, 7-day median, 9 Nov 2026).
Agnes 2.5 Pro Theta benchmarks 
GLM 4.7 Flash 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 |
|---|---|---|---|
| Coding | GLM 4.7 Flash | 55 vs 46 category score | High |
| Reasoning | GLM 4.7 Flash | 48 vs 44 category score | Medium |
| Chat quality (human preference) | – | Only GLM 4.7 Flash has an Arena rating (53 category score) | Not comparable |
| Lowest cost | Either | $0.150 vs $0.153 per 1M blended; agentic coding session $0.007 vs $0.006 | – |
| Fastest first answer | GLM 4.7 Flash | 0.3 s (GLM 4.7 Flash) vs 20.9 s (Agnes 2.5 Pro Theta) 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
- Agnes 2.5 Pro Theta and GLM 4.7 Flash are neck and neck on quality (50.8 vs 51.1, within the ±8.7 margin).
- Agnes 2.5 Pro Theta is stronger in knowledge, long context, multimodal.
- GLM 4.7 Flash is stronger in coding, reasoning, human preference, maths.
- GLM 4.7 Flash is 3.7× cheaper ($0.153 vs $0.563 per 1M blended).
- Agnes 2.5 Pro Theta streams 5.1× faster (212 vs 42 tokens per second).
At a glance
| Spec | Agnes 2.5 Pro Theta | GLM 4.7 Flash |
|---|---|---|
| Developer | Sapiens AI | Zhipu AI |
| Released | 7 Nov 2026 | 19 Jan 2026 |
| Context window (tokens) | 1,000,000 | 200,000 |
| Compared at | Agnes 2.5 Pro Theta | GLM 4.7 Flash |
| Open weights | No | Yes |
| Price in / out / cached input per 1M | $0.100 / $0.300 / – | $0.070 / $0.400 / $0.010 |
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 | Agnes 2.5 Pro Theta | GLM 4.7 Flash | Leader |
|---|---|---|---|
| Reasoning | 44.0 | 47.5 | GLM 4.7 Flash |
| Coding | 45.5 | 54.5 | GLM 4.7 Flash |
| Maths | – | 51.6 | not comparable |
| Knowledge | 38.6 | – | not comparable |
| Human preference | – | 52.9 | not comparable |
| Multimodal | 60.9 | – | not comparable |
| Long context | 67.1 | – | not comparable |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Dashed: that model has no result on the axis it crosses; the line only joins its neighbours.
Not drawn, no result for these models: Agents & tools, Instruction following.
Where sources disagree
Across the 0 benchmarks both have, Agnes 2.5 Pro Theta leads 0, GLM 4.7 Flash leads 0.
| Overall index | Agnes 2.5 Pro Theta | GLM 4.7 Flash | Leader |
|---|---|---|---|
| BenchLeader Index | 50.8 | 51.1 | neck and neck |
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. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | Agnes 2.5 Pro Theta | GLM 4.7 Flash |
|---|---|---|
| Blended price per 1M tokens | $0.150 | $0.153 |
| Input / output per 1M | $0.100 / $0.300 | $0.070 / $0.400 |
| Cached input per 1M | – | $0.010 |
| Price per task: chat reply | $0.0001 | $0.0001 |
| Price per task: summarize a 30-page report | $0.0014 | $0.0011 |
| Price per task: code edit | $0.0011 | $0.0010 |
| Price per task: agentic coding session | $0.0072 | $0.0058 |
| Price per task: structured extraction | $0.0003 | $0.0002 |
| Agentic coding session, input mostly cached | – | $0.0031 |
| Cost per benchmark task (AA Index) | $0.575 | – |
| Output tokens per task (AA Index) | 278k | – |
| Of which reasoning | 229k | – |
| Time per task (AA Index) | 22 min | – |
| Output speed | 212 tok/s | 42 tok/s |
| Time to first answer | 20.9 s | 0.3 s |
| Time to a complete answer | 29.5 s | – |
Each model is compared at its best-ranked reasoning-effort setting. Agnes 2.5 Pro Theta scores 50.8 and takes 21 s to its first answer token. GLM 4.7 Flash scores 51.1 and takes 0 s to its first answer token; GLM 4.7 Flash (no reasoning) scores 42.1 in 9 s. 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
35 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 | Agnes 2.5 Pro Theta | GLM 4.7 Flash |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 50.8 | 51.1 |
| Coding score | BenchLeader | 45.5 | 54.5 |
| Knowledge score | BenchLeader | 38.6 | – |
| Long context score | BenchLeader | 67.1 | – |
| Multimodal score | BenchLeader | 60.9 | – |
| Reasoning score | BenchLeader | 44.0 | 47.5 |
| Human preference score | BenchLeader | – | 52.9 |
| Maths score | BenchLeader | – | 51.6 |
| Blended price $/M | BenchLeader | $0.150 | $0.153 |
| Output speed | BenchLeader | 212 tok/s | 42 tok/s |
| Time to first answer | BenchLeader | 20.9 s | 0.3 s |
| Context window | BenchLeader | 1,000,000 tokens | 200,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | – | 60.5% |
| Arena Hard Prompts | Arena (LMArena) | – | 1384 |
| GPQA Diamond (AA) | Artificial Analysis | – | 45.1%† |
| Humanity's Last Exam (AA) | Artificial Analysis | 37.2% | 5.0%† |
| HLE-Diamond (AA) | Artificial Analysis | 16.8% | – |
| CritPt | Artificial Analysis | 15.1% | 0.0%† |
| CritPt v2 hard subset (AA) | Artificial Analysis | 26.0% | – |
| Coding | |||
| Arena Coding | Arena (LMArena) | – | 1423 |
| SciCode (AA) | Artificial Analysis | 56.0% | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 10.1% | – |
| Design Arena Website | Design Arena | – | 1191 |
| FrontierCode 2.0 (AA) | Artificial Analysis | 16.6% | – |
| Agents & tools | |||
| Terminal-Bench Hard | Artificial Analysis | – | 3.8%† |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 91.8%† |
| GDPval-AA v2.1 | Artificial Analysis | 1507 | – |
| AutomationBench | Artificial Analysis | 48.8% | – |
| GDP.pdf (AA) | Artificial Analysis | 14.2% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1496 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 3.8% | – |
| Maths | |||
| OTIS Mock AIME | Epoch AI Benchmarking Hub | – | 58.3% |
| Arena Maths | Arena (LMArena) | – | 1362 |
| Knowledge | |||
| AA-Omniscience v1.1 | Artificial Analysis | -11.2 | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | 21.3% | – |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | 58.6% | – |
| Instruction following | |||
| IFBench | Artificial Analysis | – | 46.3%† |
Show all 35 benchmarks
| Metric | Publisher | Agnes 2.5 Pro Theta | GLM 4.7 Flash |
|---|---|---|---|
| Instruction following | |||
| Arena Instruction Following | Arena (LMArena) | – | 1347 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1365 |
| Arena Creative Writing | Arena (LMArena) | – | 1305 |
| Arena Multi-turn | Arena (LMArena) | – | 1359 |
| Arena Longer Queries | Arena (LMArena) | – | 1370 |
| Multimodal | |||
| MMMU-Pro | Artificial Analysis | 78.4% | – |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 84.7% | 20.3%† |
| Composite | |||
| AA Intelligence Index v5.0 | Artificial Analysis | 24.0 | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | – | 10.6† |
| AA Progress Index v1 | Artificial Analysis | – | 193.5† |
Data as of 2026-11-09. Best configuration of each model; every score links to its source on the model pages.
What changed
- Agnes 2.5 Pro Theta: first results.
Related comparisons
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Agnes 2.5 Pro Theta vs GLM 4.7 Flash: questions
- Is Agnes 2.5 Pro Theta better than GLM 4.7 Flash?
- Agnes 2.5 Pro Theta and GLM 4.7 Flash are neck and neck on quality (50.8 vs 51.1, within the ±8.7 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 9 Nov 2026, so check the category scores for your use.
- Agnes 2.5 Pro Theta vs GLM 4.7 Flash: which is cheaper, and will that change?
- GLM 4.7 Flash is cheaper today: $0.153 against $0.563 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Agnes 2.5 Pro Theta or GLM 4.7 Flash better for coding?
- GLM 4.7 Flash scores higher in coding (55 vs 46 on the category index, where 50 is average).
- Which is faster, Agnes 2.5 Pro Theta or GLM 4.7 Flash?
- Agnes 2.5 Pro Theta streams 3×+ faster (212 vs 42 output tokens per second, Artificial Analysis 7-day median).
- Which is better for long documents?
- Agnes 2.5 Pro Theta accepts more context: 1,000,000 against 200,000 tokens for GLM 4.7 Flash.
More questions (2)
- Which has lower latency?
- GLM 4.7 Flash answers first: 0.3 s against 20.9 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.
- When was Agnes 2.5 Pro Theta released?
- Agnes 2.5 Pro Theta was released on 7 Nov 2026.