LLM Reference

Phi-3 Medium 128K vs Qwen3-235B-A22B

Phi-3 Medium 128K (2024) and Qwen3-235B-A22B (2025) are compact production models from Microsoft Research and Alibaba. Phi-3 Medium 128K ships a 128K-token context window, while Qwen3-235B-A22B ships a 128K-token context window. On MMLU PRO, Qwen3-235B-A22B leads by 30.9 pts. On pricing, Qwen3-235B-A22B costs $0.4/1M input tokens versus $0.5/1M for the alternative. This comparison covers specs, pricing, capabilities, benchmarks, provider availability, and production fit.

Qwen3-235B-A22B is safer overall; choose Phi-3 Medium 128K when provider fit matters.

Decision scorecard

Local evidence first
SignalPhi-3 Medium 128KQwen3-235B-A22B
Decision fitCoding, Long context, and ClassificationCoding, RAG, and Long context
Context window128K128K
Cheapest output$1.5/1M tokens$1.2/1M tokens
Provider routes2 tracked4 tracked
Shared benchmarks2 rowsMMLU PRO leader

Decision tradeoffs

Choose Phi-3 Medium 128K when...
  • Local decision data tags Phi-3 Medium 128K for Coding, Long context, and Classification.
Choose Qwen3-235B-A22B when...
  • Qwen3-235B-A22B leads the largest shared benchmark signal on MMLU PRO by 30.9 points.
  • Qwen3-235B-A22B has the lower cheapest tracked output price at $1.2/1M tokens.
  • Qwen3-235B-A22B has broader tracked provider coverage for fallback and procurement flexibility.
  • Qwen3-235B-A22B uniquely exposes Structured outputs in local model data.
  • Local decision data tags Qwen3-235B-A22B for Coding, RAG, and Long context.

Monthly cost at traffic

Estimate token spend from the cheapest tracked input and output prices on this page.

Lower estimate Qwen3-235B-A22B

Phi-3 Medium 128K

$775

Cheapest tracked route: Microsoft Foundry

Qwen3-235B-A22B

$620

Cheapest tracked route: AWS Bedrock

Estimated monthly gap: $155. Batch, cache, and negotiated pricing are excluded from this local estimate.

Switch friction

Phi-3 Medium 128K -> Qwen3-235B-A22B
  • No overlapping tracked provider route is sourced for Phi-3 Medium 128K and Qwen3-235B-A22B; plan for SDK, billing, or endpoint changes.
  • Qwen3-235B-A22B is $0.3/1M tokens lower on cheapest tracked output pricing before cache, batch, or negotiated discounts.
  • Qwen3-235B-A22B adds Structured outputs in local capability data.
Qwen3-235B-A22B -> Phi-3 Medium 128K
  • No overlapping tracked provider route is sourced for Qwen3-235B-A22B and Phi-3 Medium 128K; plan for SDK, billing, or endpoint changes.
  • Phi-3 Medium 128K is $0.3/1M tokens higher on cheapest tracked output pricing, so quality gains need to justify the spend.
  • Check replacement coverage for Structured outputs before moving production traffic.

Specs

Specification
Released2024-05-212025-04-29
Context window128K128K
Parameters14B235B
Architecturedecoder onlydecoder only
LicenseOpen SourceApache 2.0
Knowledge cutoff2023-10-

Pricing and availability

Pricing attributePhi-3 Medium 128KQwen3-235B-A22B
Input price$0.5/1M tokens$0.4/1M tokens
Output price$1.5/1M tokens$1.2/1M tokens
Providers

Capabilities

CapabilityPhi-3 Medium 128KQwen3-235B-A22B
VisionNoNo
MultimodalNoNo
ReasoningNoNo
Function callingNoNo
Tool useNoNo
Structured outputsNoYes
Code executionNoNo

Benchmarks

BenchmarkPhi-3 Medium 128KQwen3-235B-A22B
MMLU PRO51.982.8
HumanEval52.292.7

Deep dive

On shared benchmark coverage, MMLU PRO has Phi-3 Medium 128K at 51.9 and Qwen3-235B-A22B at 82.8, with Qwen3-235B-A22B ahead by 30.9 points; HumanEval has Phi-3 Medium 128K at 52.2 and Qwen3-235B-A22B at 92.7, with Qwen3-235B-A22B ahead by 40.5 points. The largest visible gap is 40.5 points on HumanEval, which matters most when that benchmark mirrors your workload. Treat isolated benchmark wins as directional, because provider routing, prompt style, and tool access can move real application results.

The capability footprint differs most on structured outputs: Qwen3-235B-A22B. Both models share the core language-model surface, so the practical split is not just feature count. Use those differences to decide whether the page is about raw model quality, agentic coding support, multimodal ingestion, or predictable structured API behavior.

For cost, Phi-3 Medium 128K lists $0.5/1M input and $1.5/1M output tokens, while Qwen3-235B-A22B lists $0.4/1M input and $1.2/1M output tokens on the cheapest tracked provider. A 70/30 input-output blend puts Qwen3-235B-A22B lower by about $0.16 per million blended tokens. Availability is 2 providers versus 4, so concentration risk also matters.

Choose Phi-3 Medium 128K when provider fit are central to the workload. Choose Qwen3-235B-A22B when provider fit, lower input-token cost, and broader provider choice are more important. For production, rerun your own prompts through the exact provider, region, and tool stack you plan to ship.

FAQ

Which has a larger context window, Phi-3 Medium 128K or Qwen3-235B-A22B?

Phi-3 Medium 128K supports 128K tokens, while Qwen3-235B-A22B supports 128K tokens. That gap matters most for long documents, large codebases, retrieval-heavy agents, and conversations where earlier context must remain visible.

Which is cheaper, Phi-3 Medium 128K or Qwen3-235B-A22B?

Qwen3-235B-A22B is cheaper on tracked token pricing. Phi-3 Medium 128K costs $0.5/1M input and $1.5/1M output tokens. Qwen3-235B-A22B costs $0.4/1M input and $1.2/1M output tokens. Provider discounts or batch pricing can still change the final bill.

Is Phi-3 Medium 128K or Qwen3-235B-A22B open source?

Phi-3 Medium 128K is listed under Open Source. Qwen3-235B-A22B is listed under Apache 2.0. License labels affect whether you can self-host, redistribute weights, or rely only on hosted APIs, so confirm the upstream license before deployment.

Which is better for structured outputs, Phi-3 Medium 128K or Qwen3-235B-A22B?

Qwen3-235B-A22B has the clearer documented structured outputs signal in this comparison. If structured outputs is mission-critical, validate it against the provider endpoint because model-level support and API-level exposure can differ.

Where can I run Phi-3 Medium 128K and Qwen3-235B-A22B?

Phi-3 Medium 128K is available on Microsoft Foundry and NVIDIA NIM. Qwen3-235B-A22B is available on Fireworks AI, AWS Bedrock, OpenRouter, and Venice AI. Provider coverage can affect latency, region availability, compliance posture, and fallback options.

When should I pick Phi-3 Medium 128K over Qwen3-235B-A22B?

Qwen3-235B-A22B is safer overall; choose Phi-3 Medium 128K when provider fit matters. If your workload also depends on provider fit, start with Phi-3 Medium 128K; if it depends on provider fit, run the same evaluation with Qwen3-235B-A22B.

Continue comparing

Last reviewed: 2026-05-20. Data sourced from public model cards and provider documentation.