GLM-4.7: benchmark scores, price and ranking

GLM-4.7 from Z.ai (Zhipu AI) ranks #410 of 743 LLMs on the LLMs Tiger aggregate benchmark score (47.7), based on 13 benchmark results captured up to 2026-10-13.

Organisation: Z.ai (Zhipu AI)
Type: Open-weight / local model
Open weights: Yes
Released: 2025-12-22 (source: epoch)
Cheapest API price: $0.400 in / $1.75 out per million tokens (via deepinfra, observed 2026-10-13T19:47Z)
Aggregate rank: #410 of 743 (score 47.7)

Relative to other models, GLM-4.7 is strongest on Epoch ECI (#35 of 204) and weakest on ProofBench (#40 of 46).

Across the 13 benchmarks where at least ten models are scored, GLM-4.7 is in the top 10% on 0 and in the top half on 6.

Its cheapest listed input price, $0.400 per million tokens, is lower than that of 53% of the 427 models with a listed price.

GLM-4.7 benchmark results

GLM-4.7 benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
Epoch ECI	144	#35 of 204	Epoch AI Benchmarking Hub	2026-10-13
GPQA Diamond	85.7%	#75 of 275	HuggingFace GPQA Diamond leaderboard	2026-10-13
OTIS Mock AIME	83.3%	#61 of 189	Epoch AI Benchmarking Hub	2026-10-13
SimpleBench	47.7%	#22 of 61	Epoch AI Benchmarking Hub	2026-10-13
SciCode	45.1%	#51 of 106	Epoch AI Benchmarking Hub	2026-10-13
WebDev Arena	1,435	#33 of 67	Epoch AI Benchmarking Hub	2026-10-13
Vending-Bench 2	$2,377	#10 of 19	Epoch AI Benchmarking Hub	2026-10-13
CritPt	1.7%	#62 of 115	Epoch AI Benchmarking Hub	2026-10-13
Chess Puzzles	6.0%	#74 of 122	Epoch AI Benchmarking Hub	2026-10-13
Terminal-Bench 2.0	33.4%	#17 of 28	Epoch AI Benchmarking Hub	2026-10-13
SimpleQA Verified	32.2%	#48 of 69	Epoch AI Benchmarking Hub	2026-10-13
ALE-Bench	399	#63 of 79	Epoch AI Benchmarking Hub	2026-10-13
ProofBench	6.0%	#40 of 46	Epoch AI Benchmarking Hub	2026-10-13

About GLM-4.7

GLM-4.7 is an open-weight model that can run on your own hardware from Z.ai (Zhipu AI). It was released on 2025-12-22, according to Epoch AI. The cheapest API listing we track charges $0.400 per million input tokens and $1.75 per million output tokens, through deepinfra. We hold 13 benchmark results for it across 3 areas, and none is in the top tenth of its board.

Where GLM-4.7 stands

On our aggregate score, GLM-4.7 is #410 of 743 ranked models, around the middle of the table. Its score is 47.7; the scale is centred on 50, so it sits 2.3 points below the typical model. That beats 45% of the ranked field. Among the 6 ranked Z.ai (Zhipu AI) models it is #6. The next one up is GLM-4.7 (Novita) at 48.4.

GLM-4.7, benchmark by benchmark

Epoch ECI puts GLM-4.7 at 144, #35 of 204. That beats 83% of the models tested, but 13 points short of Qwen 3.8 Max (156). The results either side are Qwen3-235B-A22B-Thinking (Jul 2025) (144) and Qwen 3.6 Flash (143). It covers Epoch Capabilities Index: Epoch's own aggregate of ~40 benchmarks. The board median is 127, so GLM-4.7 is 16 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

On GPQA Diamond, GLM-4.7 reaches 85.7%. It ranks #75 of 275, ahead of 73% of the field, and 10.1 points behind GPT-6 Astra (max) at 95.8%. The results either side are Claude Fable 5 (max) (85.9%) and A.X-K2 (85.6%). It covers expert-level science questions. The board median is 71.7%, so GLM-4.7 is 14.0 points above it. Source: HuggingFace GPQA Diamond leaderboard, captured 2026-10-13.

GLM-4.7 manages 83.3% on OTIS Mock AIME, good for #61 of 189. Claude Fable 5.1 (max) leads at 100.0%, 16.7 points further up, and 68% of models sit below. The results either side are o3 (high) (83.9%) and Kimi K2 Thinking Turbo (83.1%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so GLM-4.7 is 25.6 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

SimpleBench puts GLM-4.7 at 47.7%, #22 of 61. That beats 64% of the models tested, but 34.2 points short of Claude Fable 5 (max) (81.9%). The results either side are Gemini 2.5 Pro Preview (Mar 2025) (51.6%) and GLM-4.7 (Novita) (47.7%). It covers trick questions that test common sense. The board median is 41.4%, so GLM-4.7 is 6.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

On SciCode, GLM-4.7 reaches 45.1%. It ranks #51 of 106, ahead of 52% of the field, and 21.8 points behind Claude Opus 5.5 (max) at 66.9%. The results either side are Qwen3.7 Plus (45.5%) and DeepSeek v4 Flash (max) (44.9%). It covers scientific research code. The board median is 44.3%, so GLM-4.7 is 0.8 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

GLM-4.7 manages 1,435 on WebDev Arena, good for #33 of 67. Claude Opus 5.5 (max) leads at 1,820, 386 points further up, and 51% of models sit below. The results either side are GLM-5 (1,436) and mimo-v2-pro (1,433). It covers building web apps, crowd-voted (Elo). The board median is 1,433, so GLM-4.7 is 1 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

GLM-4.7 lands below the middle on Terminal-Bench 2.0: 33.4%, #17 of 28, with 57% of the field scoring higher. Gemini 3 Pro Preview leads at 69.4%; the closest result above is Kimi K2 Thinking at 35.7%. It covers real terminal tasks (agent + model). The board median is 37.3%, so GLM-4.7 is 3.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

GLM-4.7 lands below the middle on SimpleQA Verified: 32.2%, #48 of 69, with 68% of the field scoring higher. gemini-4-argon-high leads at 76.0%; the closest result above is Claude Sonnet 4.6 (max) at 32.8%. It covers short factual questions. The board median is 41.0%, so GLM-4.7 is 8.8 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

GLM-4.7 lands below the middle on ALE-Bench: 399, #63 of 79, with 78% of the field scoring higher. GPT-6 Astra (max) leads at 2,951; the closest result above is Ring-2.6-1T at 433. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so GLM-4.7 is 364 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

ProofBench is a weak spot. GLM-4.7 scores 6.0%, #40 of 46, behind 85% of the models measured. The leader, Claude Fable 5.1 (max), reaches 100.0%. It covers writing mathematical proofs. The board median is 21.0%, so GLM-4.7 is 15.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

The full table above also lists the other 3 results, including boards with fewer than ten models, where a rank says little.

Strengths and weaknesses

Software engineering: uneven, ahead of 41% of the field on average across 4 benchmarks. Its best board here is SciCode (#51 of 106) and its weakest is ALE-Bench (#63 of 79). Reasoning, maths and science: solid, ahead of 53% of the field on average across 5 benchmarks. Its best board here is GPQA Diamond (#75 of 275) and its weakest is ProofBench (#40 of 46). General knowledge and chat: uneven, ahead of 50% of the field on average across 4 benchmarks. Its best board here is Epoch ECI (#35 of 204) and its weakest is SimpleQA Verified (#48 of 69).

Price and value

The median input price across the 427 priced models is $0.500 per million tokens, so GLM-4.7 costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on Kimi K2.5 (instant) ($0.450) and gpt-3.5-turbo-0125 ($0.500).

Release history

The next one was GLM-4.7 (Novita) on 2025-12-22, at 48.4.

Reading these numbers

The 13 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (12), HuggingFace GPQA Diamond leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements.

Models ranked nearby

#407 Grok 4 (Fast Reasoning) (47.7)
#408 Mistral-Small-24B-Base-2501 (47.7)
#409 Gemini 2.5 Pro (Thinking 32K) (47.7)
#411 Mistral Large 4 (high) (47.7)
#412 Claude Haiku 4.5 (Thinking 32K) (47.7)
#413 gemini-3.1-flash-lite (high) (47.7)