ZMIME
Comparison · 3 models · Updated Nov 22, 2026

Aya Expanse 32B vs Phi-4-mini vs Qwen2.5-Coder-0.5B

No independent capability benchmark covers every model here yet. The score uses context window.

  1. Cohere

    Aya Expanse 32B

    Released Oct 24, 2024

    32/100
    • ECI—
    • Price$0.50 / $1.50
    • Context128K
  2. Microsoft

    Phi-4-mini

    Released Feb 26, 2025

    32/100
    • ECI—
    • Price$0.075 / $0.30
    • Context128K
  3. Alibaba (Qwen)

    Qwen2.5-Coder-0.5B

    Released Nov 12, 2024

    11/100
    • ECI68.1
    • Price—
    • Context33K
  4. Cohere

    Aya Expanse 32B

    Released Oct 24, 2024

    32/100
    • ECI—
    • Price$0.50 / $1.50
    • Context128K
  5. Microsoft

    Phi-4-mini

    Released Feb 26, 2025

    32/100
    • ECI—
    • Price$0.075 / $0.30
    • Context128K
  6. Alibaba (Qwen)

    Qwen2.5-Coder-0.5B

    Released Nov 12, 2024

    11/100
    • ECI68.1
    • Price—
    • Context33K
01 — Verdict

Capability evidence is incomplete

There is not enough shared capability evidence to name an overall winner. The scores (Aya Expanse 32B 32/100, Phi-4-mini 32/100, Qwen2.5-Coder-0.5B 11/100) compare the available specifications and price. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.There is not enough shared capability evidence to name an overall winner. The scores (Aya Expanse 32B 32/100, Phi-4-mini 32/100, Qwen2.5-Coder-0.5B 11/100) compare the available specifications only. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

  • CapabilityNot enough dataNo independent benchmark covers every model here yet
  • Lowest pricePhi-4-miniPhi-4-mini $0.131 · Aya Expanse 32B $0.75 per 1M tokens (3:1 blend) · Qwen2.5-Coder-0.5B unpriced
  • Longest contextAya Expanse 32B and Phi-4-miniAya Expanse 32B 128,000 · Phi-4-mini 128,000 · Qwen2.5-Coder-0.5B 32,768 tokens
  • Widest inputsSame inputsAya Expanse 32B: Text · Phi-4-mini: Text · Qwen2.5-Coder-0.5B: Text
  • Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
How the score is built
MeasureWeightAya Expanse 32BPhi-4-miniQwen2.5-Coder-0.5B
Context window100%323211
Specs & price only100%32/10032/10011/100

Left out because at least one model lacks the data: capability, price and inputs & features. The remaining weights were rescaled.

How the score is built
MeasureWeightAya Expanse 32BPhi-4-miniQwen2.5-Coder-0.5B
Context window100%323211
Specs only100%32/10032/10011/100

Left out because at least one model lacks the data: capability and inputs & features. The remaining weights were rescaled.

02 — Side by side

Every spec in one table

Highlighted cells lead their row. Unavailable specifications are marked “Not published”. Publisher results include their evaluation settings and source.

Aya Expanse 32B vs Phi-4-mini vs Qwen2.5-Coder-0.5B specifications side by side
SpecificationAya Expanse 32BCoherePhi-4-miniMicrosoftQwen2.5-Coder-0.5BAlibaba (Qwen)
Capability
Capabilities Index (ECI)Not publishedNot published68.1
ECI rankNot publishedNot published#159 of 159
ECI confidence intervalNot publishedNot published48.4–83.9
ECI evaluation dateNot publishedNot publishedSep 18, 2024
Documented capabilitiesOpen multilingual model optimized for generation across 23 languagesCompact Microsoft instruction model tuned for efficient coding assistance, reasoning, and low-latency agent tasksTiny open Qwen code model for lightweight completion and on-device coding
Model parametersNot publishedNot publishedNot published
Output modalitiesTextTextText
Publisher-reported benchmarks
MMLUPublisher-reported resultNot published67.3accuracyReported resultNot published
Price per million tokens
Input$0.50$0.075 (best)Not published
Output$1.50$0.30 (best)Not published
Cached inputNot publishedNot publishedNot published
Blended (3:1)$0.75$0.131 (best)Not published
Long-context rateSame rateSame rateNot published
Price sourceOfficial Cohere APIOfficial Azure APINot published
Limits
Context window128,000 tokens (best)128,000 tokens (best)32,768 tokens
Max inputNot publishedNot publishedNot published
Max output4,000 tokens4,096 tokens8,192 tokens (best)
Inputs and features
TextYesYesYes
ImagesNoNoNo
PDFsNoNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningNoNoNo
Tool callingNoYesNo
Structured outputNoNot publishedNot published
File attachmentsNoNoNo
Temperature controlYesYesYes
Reasoning controlNot publishedNot publishedNot published
Availability
WeightsOpenCC-BY-NC-4.0OpenOpenApache 2.0
API model IDc4ai-aya-expanse-32bphi-4-miniNot published
API providers22Not published
ReleasedOct 24, 2024Feb 26, 2025Nov 12, 2024
Knowledge cutoffNot publishedJun 2024Not published
Publisher documentationPublisher documentation · Model weightsPublisher documentation · Model weightsModel weights
Source metadata updatedOct 24, 2024Feb 26, 2025Nov 12, 2024
Data sourcesmodels.dev · Epoch AImodels.dev · Epoch AImodels.dev · Epoch AI
Catalog checkedNov 22, 2026Nov 22, 2026Nov 22, 2026
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.

  • Aya Expanse 32B$8.00
  • Phi-4-mini$1.35
  • Qwen2.5-Coder-0.5B—
04 — Questions

Which should you choose?

Which is better: Aya Expanse 32B, Phi-4-mini or Qwen2.5-Coder-0.5B?

There is not enough shared capability evidence to name an overall winner. The scores (Aya Expanse 32B 32/100, Phi-4-mini 32/100, Qwen2.5-Coder-0.5B 11/100) compare the available specifications and price. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.There is not enough shared capability evidence to name an overall winner. The scores (Aya Expanse 32B 32/100, Phi-4-mini 32/100, Qwen2.5-Coder-0.5B 11/100) compare the available specifications only. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

Which is cheaper, Aya Expanse 32B, Phi-4-mini or Qwen2.5-Coder-0.5B?

Phi-4-mini is cheaper at $0.075 input / $0.30 output per million tokens (official Azure API price). Aya Expanse 32B costs $0.50 input / $1.50 output per million tokens (official Cohere API price). At a typical mix of three input tokens to one output token, that is $0.131 per million tokens for Phi-4-mini versus $0.75 for Aya Expanse 32B (5.7× as much). Qwen2.5-Coder-0.5B has no published per-token price.

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Aya Expanse 32B has not been scored yet, Phi-4-mini has not been scored yet and Qwen2.5-Coder-0.5B has an ECI of 68.1.

Which is better for coding?

There are no published SWE-bench Verified results for Aya Expanse 32B, Phi-4-mini and Qwen2.5-Coder-0.5B yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Aya Expanse 32B and Qwen2.5-Coder-0.5B has no confirmed tool-calling support in this comparison.

Which has the bigger context window?

Aya Expanse 32B and Phi-4-mini have the largest context windows (128,000 and 128,000 tokens), against 32,768 for Qwen2.5-Coder-0.5B. Maximum output per response: Aya Expanse 32B up to 4,000, Phi-4-mini up to 4,096, Qwen2.5-Coder-0.5B up to 8,192 tokens.

Which can read images, PDFs, audio or video?

Aya Expanse 32B accepts text; Phi-4-mini accepts text; Qwen2.5-Coder-0.5B accepts text. They handle the same number of input types.

Are any of these open source?

Yes, all three publish their weights (CC-BY-NC-4.0 and Apache 2.0), so you can self-host them.

Which is newer?

Phi-4-mini is the newest, released Feb 26, 2025. Qwen2.5-Coder-0.5B came out Nov 12, 2024; Aya Expanse 32B came out Oct 24, 2024. Knowledge cutoff: Phi-4-mini Jun 2024.

How do you decide the winner?

This rating uses context window (100%, 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 22, 2026.