Skip to main content
Radar

Provider changes are easy to miss. Radar watches releases, pricing, deprecations, and incidents at the source.Provider changes are easy to miss.

See Radar
BenchLM
Advertise

Model comparison

Composer 2 vs GLM-5

Updated August 7, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Composer 2

Cursor

—

Evidence status unavailable

90% interval unavailable

GLM-5

Z.AI

65.3/100

Supported · Public rank #32

90% interval 54.5–76.1

At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.

4 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

  • Agentic work

    Tool use, computer use, and multi-step task completion

    Composer 2

    Composer 2 leads on the same 1 weighted benchmark row.

    Confidence: limited

  • Chat turn cost

    1K fresh input + 500 output tokens

    Composer 2

    Composer 2 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

Show secondary and unsupported calls

  • Repository review cost

    50K fresh input + 3K output tokens

    Composer 2

    Composer 2 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

  • Coding work

    Code generation, repair, and software-engineering tasks

    Not enough matched evidence

    The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.

    Confidence: limited

  • Long documents

    Prompts that approach the documented context limit

    No clear pick

    The documented context windows are equal.

    Confidence: documented

  • Cache-heavy agent loop cost

    200K cached + 20K fresh input + 10K output tokens

    Not enough matched evidence

    The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Composer 2 does not fit this workload in one request. GLM-5 does not fit this workload in one request. Composer 2 has no published cached-input rate, so cached tokens use its listed input rate. GLM-5 has no published cached-input rate, so cached tokens use its listed input rate.

    Confidence: rate-fallback

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

Shared results
4
Composer 2 only
1
GLM-5 only
32
Like-for-like categories
1 / 8

1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.

Agentic

Like-for-like
Composer 2
61.7
GLM-5
56.2
Weighted basis
1 vs 1 rows
Reading
Composer 2 leads

Coding

Directional only
Composer 2
58.0
GLM-5
66.3
Weighted basis
1 vs 3 rows
Reading
Directional only

Reasoning

Not comparable
Composer 2
Not measured
GLM-5
60.8
Weighted basis
0 vs 1 rows
Reading
Not comparable

Knowledge

Not comparable
Composer 2
Not measured
GLM-5
66.4
Weighted basis
0 vs 4 rows
Reading
Not comparable

Math

Not comparable
Composer 2
Not measured
GLM-5
56.3
Weighted basis
0 vs 4 rows
Reading
Not comparable

Multilingual

Not comparable
Composer 2
Not measured
GLM-5
83.1
Weighted basis
0 vs 1 rows
Reading
Not comparable

Multimodal

Not comparable
Composer 2
Not measured
GLM-5
Not measured
Weighted basis
0 vs 0 rows
Reading
Not comparable

Instruction following

Not comparable
Composer 2
Not measured
GLM-5
92.6
Weighted basis
0 vs 1 rows
Reading
Not comparable

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

Composer 2
$0.00175
Fits in one request
GLM-5
$0.0026
Fits in one request

Composer 2 has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Composer 2
$0.0325
Fits in one request
GLM-5
$0.0596
Fits in one request

Composer 2 has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Composer 2
$0.135
Does not fit in one request
Cached input priced at the published list-input rate
GLM-5
$0.252
Does not fit in one request
Cached input priced at the published list-input rate

Composer 2 does not fit this workload in one request. GLM-5 does not fit this workload in one request. Composer 2 has no published cached-input rate, so cached tokens use its listed input rate. GLM-5 has no published cached-input rate, so cached tokens use its listed input rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

Context window

Maximum documented context; output-token limits may be lower.

Composer 2

200K

GLM-5

200K

API model ID

Composer 2

Not sourced

GLM-5

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Composer 2

Not published

GLM-5

Not published

Documented inputs

Composer 2

Not sourced

GLM-5

Not sourced

Documented outputs

Composer 2

Not sourced

GLM-5

Not sourced

Provider availability

Composer 2

Not sourced

GLM-5

Not sourced

Reasoning profile

Composer 2

Reasoning

GLM-5

Non-Reasoning

Weight access

Composer 2

Proprietary

GLM-5

Open Weight

License

Composer 2

Proprietary

GLM-5

Open Weight

Release date

Composer 2

2026-03-19

GLM-5

2026-03-01

If you already use one of these models
Deployment change
The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
Workload cost
Repository review: $0.0325 vs $0.0596. Cache-heavy agent loop: $0.135 vs $0.252.
Context tradeoff
Both models list 200K.

Run the same representative tasks against both endpoints before changing production traffic.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence37 rows

Agentic

  • Terminal-Bench 2.0

    Composer 261.7%
    Source
    GLM-556.2%
    Source

    Composer 2 leads this result

  • Claw-Eval

    Composer 2—
    GLM-557.7%
    Source

    Not directly comparable

  • QwenClawBench

    Composer 2—
    GLM-554.1%
    Source

    Not directly comparable

  • τ³-bench results

    Composer 2—
    GLM-565.6%
    Source

    Not directly comparable

  • DeepPlanning

    Composer 2—
    GLM-514.6%
    Source

    Not directly comparable

  • Toolathlon

    Composer 2—
    GLM-538%
    Source

    Not directly comparable

  • MCP Atlas

    Composer 2—
    GLM-531.1%
    Source

    Not directly comparable

  • MCP-Tasks

    Composer 2—
    GLM-560.8%
    Source

    Not directly comparable

  • WideResearch

    Composer 2—
    GLM-569.8%
    Source

    Not directly comparable

  • CyberGym

    Composer 2—
    GLM-543.2%
    Source

    Not directly comparable

  • Gert Labs

    Composer 2—
    GLM-550.99%
    Source

    Not directly comparable

Coding

  • SWE Multilingual

    Composer 273.7%
    Source
    GLM-573.3%
    Source

    Composer 2 leads this result

  • SWE-Rebench

    Composer 258%
    Source
    GLM-562.8%
    Source

    GLM-5 leads this result

  • React Native Evals

    Shared source
    Composer 296.1%
    GLM-574.8%

    Composer 2 leads this result

  • Terminal-Bench 2.0

    Composer 261.7%
    Source
    GLM-5—

    Not directly comparable

  • SWE-bench Verified

    Composer 2—
    GLM-577.8%
    Source

    Not directly comparable

  • SWE-bench Verified*

    Composer 2—
    GLM-572.8%
    Source

    Not directly comparable

  • SWE-bench Pro

    Composer 2—
    GLM-555.1%
    Source

    Not directly comparable

Reasoning

  • LongBench v2

    Composer 2—
    GLM-560.8%
    Source

    Not directly comparable

  • AI-Needle

    Composer 2—
    GLM-563.3%
    Source

    Not directly comparable

Knowledge

  • GPQA

    Composer 2—
    GLM-586%
    Source

    Not directly comparable

  • GPQA-D

    Composer 2—
    GLM-586.0%
    Source

    Not directly comparable

  • SuperGPQA

    Composer 2—
    GLM-566.8%
    Source

    Not directly comparable

  • MMLU-Pro

    Composer 2—
    GLM-585.7%
    Source

    Not directly comparable

  • MMLU-Pro (Arcee)

    Composer 2—
    GLM-585.8%
    Source

    Not directly comparable

  • HLE

    Composer 2—
    GLM-550.4%
    Source

    Not directly comparable

Math

  • AIME26

    Composer 2—
    GLM-595.8%
    Source

    Not directly comparable

  • AIME25 (Arcee)

    Composer 2—
    GLM-593.3%
    Source

    Not directly comparable

  • HMMT Feb 2025

    Composer 2—
    GLM-597.5%
    Source

    Not directly comparable

  • HMMT Nov 2025

    Composer 2—
    GLM-596.9%
    Source

    Not directly comparable

  • HMMT Feb 2026

    Composer 2—
    GLM-586.4%
    Source

    Not directly comparable

  • MMAnswerBench

    Composer 2—
    GLM-582.5%
    Source

    Not directly comparable

  • FrontierMath v2 (Tiers 1-3)

    Composer 2—
    GLM-516.434%
    Source

    Not directly comparable

  • FrontierMath v2 (Tier 4)

    Composer 2—
    GLM-52.100%
    Source

    Not directly comparable

Multilingual

  • MMLU-ProX

    Composer 2—
    GLM-583.1%
    Source

    Not directly comparable

  • NOVA-63

    Composer 2—
    GLM-555.1%
    Source

    Not directly comparable

Instruction following

  • IFEval

    Composer 2—
    GLM-592.6%
    Source

    Not directly comparable

Frequently asked questions

Which is better, Composer 2 or GLM-5?

At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.

Which is better for coding, Composer 2 or GLM-5?

The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, Composer 2 or GLM-5?

Composer 2 leads the like-for-like agentic tasks comparison across 1 shared weighted benchmark row.

Which costs less, Composer 2 or GLM-5?

For the stated presets, chat costs $0.00175 on Composer 2 and $0.0026 on GLM-5; repository review costs $0.0325 and $0.0596; the cache-heavy agent loop costs $0.135 and $0.252. Composer 2 does not fit this workload in one request. GLM-5 does not fit this workload in one request. Composer 2 has no published cached-input rate, so cached tokens use its listed input rate. GLM-5 has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Composer 2 or GLM-5?

Both models list the same context window, 200K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 7, 2026

Watch Composer 2 vs GLM-5

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.