ZMIME
Comparison · 3 models · Updated Nov 2, 2026

Mixtral 8x22B Instruct vs Claude Haiku 3 vs Gemma 2 27B

Claude Haiku 3 comes out ahead, 46 to 35 and 34 on our weighted score, and it is the cheaper option too.

  1. Mistral AI

    Mixtral 8x22B Instruct

    Released Apr 17, 2024

    34/100
    • ECI122.0
    • Price$2.00 / $6.00
    • Context66K
    Full specs Remove Mixtral 8x22B Instruct
  2. Our pick

    Anthropic

    Claude Haiku 3

    Released Mar 13, 2024

    46/100
    • ECI118.8
    • Price$0.25 / $1.25
    • Context200K
    Full specs Remove Claude Haiku 3

  3. Google

    Gemma 2 27B

    Released Jul 13, 2024

    35/100
    • ECI122.1
    • Price$0.65 / $0.65
    • Context8K
    Full specs Remove Gemma 2 27B
01 — Verdict

Claude Haiku 3 is our pick

Claude Haiku 3 is the better all-round choice, scoring 46/100 against Gemma 2 27B (35) and Mixtral 8x22B Instruct (34). It leads on price, inputs & features and context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

  • CapabilityGemma 2 27BCapabilities Index (ECI): Gemma 2 27B 122.1 · Mixtral 8x22B Instruct 122.0 · Claude Haiku 3 118.8
  • Lowest priceClaude Haiku 3Claude Haiku 3 $0.50 · Gemma 2 27B $0.65 · Mixtral 8x22B Instruct $3.00 per 1M tokens (3:1 blend)
  • Longest contextClaude Haiku 3Claude Haiku 3 200,000 · Mixtral 8x22B Instruct 65,536 · Gemma 2 27B 8,192 tokens
  • Widest inputsClaude Haiku 3Mixtral 8x22B Instruct: Text, PDFs · Claude Haiku 3: Text, Images, PDFs · Gemma 2 27B: Text
  • Self-hostingMixtral 8x22B Instruct and Gemma 2 27BPublishes downloadable weights
How the score is built
MeasureWeightMixtral 8x22B InstructClaude Haiku 3Gemma 2 27B
CapabilityCapabilities Index (ECI)50%423842
Price25%266156
Inputs & features15%35600
Context window10%12320
Overall100%34/10046/10035/100
02 — Side by side

Every spec in one table

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

Mixtral 8x22B Instruct vs Claude Haiku 3 vs Gemma 2 27B specifications side by side
SpecificationMixtral 8x22B InstructMistral AIClaude Haiku 3AnthropicGemma 2 27BGoogle
Capability
Capabilities Index (ECI)122.0118.8122.1 (best)
ECI rank#155 of 165#156 of 165#152 of 165 (best)
GPQA DiamondGraduate-level science questions34.1%36.3%36.5% (best)
OTIS Mock AIME 2024–2025Competition mathematics—1.8% (best)1.4%
Price per million tokens
Input$2.00$0.25 (best)$0.65
Output$6.00$1.25$0.65 (best)
Cached input———
Blended (3:1)$3.00$0.50 (best)$0.65
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 5 providersMedian of 3 providersMedian of 3 providers
Limits
Context window65,536 tokens200,000 tokens (best)8,192 tokens
Max output52,428 tokens (best)4,096 tokens2,048 tokens
Inputs and features
TextYesYesYes
ImagesNoYesNo
PDFsYesYesNo
AudioNoNoNo
VideoNoNoNo
ReasoningNoNoNo
Tool callingYesYesNo
Structured outputNoNoNo
Availability
WeightsOpenProprietaryOpen
API model ID———
API providers5 (best)33
ReleasedApr 17, 2024Mar 13, 2024Jul 13, 2024
Knowledge cutoffJan 31, 2024Aug 31, 2023Jun 30, 2024
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.

  • Mixtral 8x22B Instruct$32.00
  • Claude Haiku 3$5.00
  • Gemma 2 27B$7.80
04 — Questions

Which should you choose?

Which is better: Mixtral 8x22B Instruct, Claude Haiku 3 or Gemma 2 27B?

Claude Haiku 3 is the better all-round choice, scoring 46/100 against Gemma 2 27B (35) and Mixtral 8x22B Instruct (34). It leads on price, inputs & features and context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Which is cheaper, Mixtral 8x22B Instruct, Claude Haiku 3 or Gemma 2 27B?

Claude Haiku 3 is cheaper at $0.25 input / $1.25 output per million tokens (median across 3 API providers). Gemma 2 27B costs $0.65 input / $0.65 output per million tokens (median across 3 API providers); Mixtral 8x22B Instruct costs $2.00 input / $6.00 output per million tokens (median across 5 API providers). At a typical mix of three input tokens to one output token, that is $0.50 per million tokens for Claude Haiku 3 versus $0.65 for Gemma 2 27B (1.3× as much) and $3.00 for Mixtral 8x22B Instruct (6× as much).

Which scores higher on benchmarks?

Gemma 2 27B scores higher on the Capabilities Index (ECI): Gemma 2 27B 122.1 (#152 of 165), Mixtral 8x22B Instruct 122.0 (#155 of 165) and Claude Haiku 3 118.8 (#156 of 165). The confidence ranges of the top two overlap (115.8–124.1 vs 115.7–124.3), so treat the gap as small. On individual benchmarks: GPQA Diamond — Gemma 2 27B 36.5%, Claude Haiku 3 36.3%, Mixtral 8x22B Instruct 34.1%.

Which is better for coding?

There are no published SWE-bench Verified results for Mixtral 8x22B Instruct, Claude Haiku 3 and Gemma 2 27B yet, so there is no like-for-like coding score. On overall capability, Gemma 2 27B leads, which tends to carry over to coding, but test on your own codebase. Note that Gemma 2 27B does not support tool calling, which most coding agents need.

Which has the bigger context window?

Claude Haiku 3 has the largest context window at 200,000 tokens, against 65,536 for Mixtral 8x22B Instruct and 8,192 for Gemma 2 27B. Maximum output per response: Mixtral 8x22B Instruct up to 52,428, Claude Haiku 3 up to 4,096, Gemma 2 27B up to 2,048 tokens.

Which can read images, PDFs, audio or video?

Mixtral 8x22B Instruct accepts text and PDFs; Claude Haiku 3 accepts text, images and PDFs; Gemma 2 27B accepts text. Claude Haiku 3 handles the widest range of inputs.

Are any of these open source?

Mixtral 8x22B Instruct and Gemma 2 27B publishes its weights and can be self-hosted; Claude Haiku 3 is proprietary.

Which is newer?

Gemma 2 27B is the newest, released Jul 13, 2024. Mixtral 8x22B Instruct came out Apr 17, 2024; Claude Haiku 3 came out Mar 13, 2024. Knowledge cutoff: Mixtral 8x22B Instruct Jan 31, 2024, Claude Haiku 3 Aug 31, 2023, Gemma 2 27B Jun 30, 2024.

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