ZMIME
Comparison · 3 models · Updated Nov 2, 2026

Qwen3 235B-A22B Thinking 2507 vs GPT OSS 20B vs GLM-4.7

GPT OSS 20B comes out ahead, 61 to 55 and 55 on our weighted score, and it is the cheaper option too.

  1. Alibaba (Qwen)

    Qwen3 235B-A22B Thinking 2507

    Released Jul 25, 2025

    55/100
    • ECI143.8
    • Price$0.30 / $2.90
    • Context262K
    Full specs Remove Qwen3 235B-A22B Thinking 2507
  2. Our pick

    OpenAI

    GPT OSS 20B

    Released Aug 5, 2025

    61/100
    • ECI137.8
    • Price$0.07 / $0.25
    • Context131K
    Full specs Remove GPT OSS 20B

  3. Z.ai (Zhipu)

    GLM-4.7

    Released Dec 22, 2025

    55/100
    • ECI143.7
    • Price$0.60 / $2.20
    • Context205K
    Full specs Remove GLM-4.7
01 — Verdict

GPT OSS 20B is our pick

GPT OSS 20B is the better all-round choice, scoring 61/100 against Qwen3 235B-A22B Thinking 2507 (55) and GLM-4.7 (55). It leads on price. Qwen3 235B-A22B Thinking 2507 wins on context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

  • CapabilityQwen3 235B-A22B Thinking 2507Capabilities Index (ECI): Qwen3 235B-A22B Thinking 2507 143.8 · GLM-4.7 143.7 · GPT OSS 20B 137.8
  • Lowest priceGPT OSS 20BGPT OSS 20B $0.115 · Qwen3 235B-A22B Thinking 2507 $0.95 · GLM-4.7 $1.00 per 1M tokens (3:1 blend)
  • Longest contextQwen3 235B-A22B Thinking 2507Qwen3 235B-A22B Thinking 2507 262,144 · GLM-4.7 204,800 · GPT OSS 20B 131,072 tokens
  • Widest inputsSame inputsQwen3 235B-A22B Thinking 2507: Text · GPT OSS 20B: Text · GLM-4.7: Text
  • Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
How the score is built
MeasureWeightQwen3 235B-A22B Thinking 2507GPT OSS 20BGLM-4.7
CapabilityCapabilities Index (ECI)50%696169
Price25%499048
Inputs & features15%353535
Context window10%372532
Overall100%55/10061/10055/100
02 — Side by side

Every spec in one table

Highlighted cells lead their row. Dashes mean the data is not published.

Qwen3 235B-A22B Thinking 2507 vs GPT OSS 20B vs GLM-4.7 specifications side by side
SpecificationQwen3 235B-A22B Thinking 2507Alibaba (Qwen)GPT OSS 20BOpenAIGLM-4.7Z.ai (Zhipu)
Capability
Capabilities Index (ECI)143.8 (best)137.8143.7
ECI rank#93 of 165 (best)#122 of 165#94 of 165
GPQA DiamondGraduate-level science questions80.1%60.8%83.3% (best)
OTIS Mock AIME 2024–2025Competition mathematics86.7% (best)65.3%83.3%
SimpleQA VerifiedShort factual questions40.4% (best)—32.2%
Price per million tokens
Input$0.30$0.07 (best)$0.60
Output$2.90$0.25 (best)$2.20
Cached input——$0.11
Blended (3:1)$0.95$0.115 (best)$1.00
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 11 providersMedian of 37 providersOfficial Z.AI API
Limits
Context window262,144 tokens (best)131,072 tokens204,800 tokens
Max output81,920 tokens32,768 tokens131,072 tokens (best)
Inputs and features
TextYesYesYes
ImagesNoNoNo
PDFsNoNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningYesYesYes
Tool callingYesYesYes
Structured outputNoNoNo
Availability
WeightsOpenApache 2.0OpenOpen
API model ID——glm-4.7
API providers1241 (best)31
ReleasedJul 25, 2025Aug 5, 2025Dec 22, 2025
Knowledge cutoff—Jun 2024Apr 2025
03 — Cost

What would a month cost?

Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.

  • Qwen3 235B-A22B Thinking 2507$8.80
  • GPT OSS 20B$1.20
  • GLM-4.7$10.40
04 — Questions

Which should you choose?

Which is better: Qwen3 235B-A22B Thinking 2507, GPT OSS 20B or GLM-4.7?

GPT OSS 20B is the better all-round choice, scoring 61/100 against Qwen3 235B-A22B Thinking 2507 (55) and GLM-4.7 (55). It leads on price. Qwen3 235B-A22B Thinking 2507 wins on context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Which is cheaper, Qwen3 235B-A22B Thinking 2507, GPT OSS 20B or GLM-4.7?

GPT OSS 20B is cheaper at $0.07 input / $0.25 output per million tokens (median across 37 API providers). Qwen3 235B-A22B Thinking 2507 costs $0.30 input / $2.90 output per million tokens (median across 11 API providers); GLM-4.7 costs $0.60 input / $2.20 output per million tokens (official Z.AI API price). At a typical mix of three input tokens to one output token, that is $0.115 per million tokens for GPT OSS 20B versus $0.95 for Qwen3 235B-A22B Thinking 2507 (8.3× as much) and $1.00 for GLM-4.7 (8.7× as much).

Which scores higher on benchmarks?

Qwen3 235B-A22B Thinking 2507 scores higher on the Capabilities Index (ECI): Qwen3 235B-A22B Thinking 2507 143.8 (#93 of 165), GLM-4.7 143.7 (#94 of 165) and GPT OSS 20B 137.8 (#122 of 165). The confidence ranges of the top two overlap (141.9–145.2 vs 141.4–145.8), so treat the gap as small. On individual benchmarks: GPQA Diamond — GLM-4.7 83.3%, Qwen3 235B-A22B Thinking 2507 80.1%, GPT OSS 20B 60.8%; OTIS Mock AIME 2024–2025 — Qwen3 235B-A22B Thinking 2507 86.7%, GLM-4.7 83.3%, GPT OSS 20B 65.3%.

Which is better for coding?

There are no published SWE-bench Verified results for Qwen3 235B-A22B Thinking 2507, GPT OSS 20B and GLM-4.7 yet, so there is no like-for-like coding score. On overall capability, Qwen3 235B-A22B Thinking 2507 leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.

Which has the bigger context window?

Qwen3 235B-A22B Thinking 2507 has the largest context window at 262,144 tokens, against 204,800 for GLM-4.7 and 131,072 for GPT OSS 20B. Maximum output per response: Qwen3 235B-A22B Thinking 2507 up to 81,920, GPT OSS 20B up to 32,768, GLM-4.7 up to 131,072 tokens.

Which can read images, PDFs, audio or video?

Qwen3 235B-A22B Thinking 2507 accepts text; GPT OSS 20B accepts text; GLM-4.7 accepts text. They handle the same number of input types.

Are any of these open source?

Yes, all three publish their weights (Apache 2.0), so you can self-host them.

Which is newer?

GLM-4.7 is the newest, released Dec 22, 2025. GPT OSS 20B came out Aug 5, 2025; Qwen3 235B-A22B Thinking 2507 came out Jul 25, 2025. Knowledge cutoff: GPT OSS 20B Jun 2024, GLM-4.7 Apr 2025.

How do you decide the winner?

Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 2, 2026.

05 — Keep comparing

Related comparisons