ZMIME
Comparison · 3 models · Updated Nov 18, 2026

Llama-3.3-70B-Instruct vs Mistral Small 3.1 24B vs Nova Lite

No independent capability benchmark covers every model here yet. The score uses inputs & features and context window.

Price in rating: Excluded · Include price

  1. Meta

    Llama-3.3-70B-Instruct

    Released Dec 6, 2024

    28/100
    • ECI126.2
    • Price$0.59 / $0.745
    • Context128K
  2. Mistral AI

    Mistral Small 3.1 24B

    Released Mar 17, 2025

    43/100
    • ECI—
    • Price$0.351 / $0.555
    • Context128K
  3. Amazon

    Nova Lite

    Released Dec 3, 2024

    60/100
    • ECI—
    • Price$0.06 / $0.24
    • Context300K
01 — Verdict

Capability evidence is incomplete

There is not enough shared capability evidence to name an overall winner. The scores (Nova Lite 60/100, Mistral Small 3.1 24B 43/100, Llama-3.3-70B-Instruct 28/100) compare the available specifications only. The score weighs inputs & features 60%, context window 40%. 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 priceNova LiteNova Lite $0.105 · Mistral Small 3.1 24B $0.402 · Llama-3.3-70B-Instruct $0.629 per 1M tokens (3:1 blend)
  • Longest contextNova LiteNova Lite 300,000 · Llama-3.3-70B-Instruct 128,000 · Mistral Small 3.1 24B 128,000 tokens
  • Widest inputsNova LiteLlama-3.3-70B-Instruct: Text · Mistral Small 3.1 24B: Text, Images · Nova Lite: Text, Images, PDFs, Video
  • Self-hostingLlama-3.3-70B-Instruct and Mistral Small 3.1 24BPublishes downloadable weights
How the score is built
MeasureWeightLlama-3.3-70B-InstructMistral Small 3.1 24BNova Lite
Inputs & features60%255070
Context window40%323245
Specs only100%28/10043/10060/100

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

02 — Side by side

Every spec in one table

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

Llama-3.3-70B-Instruct vs Mistral Small 3.1 24B vs Nova Lite specifications side by side
SpecificationLlama-3.3-70B-InstructMetaMistral Small 3.1 24BMistral AINova LiteAmazon
Capability
Capabilities Index (ECI)126.2——
ECI rank#144 of 158——
GPQA DiamondGraduate-level science questions47.4%47.5% (best)—
OTIS Mock AIME 2024–2025Competition mathematics5.1%5.8% (best)—
Price per million tokens
Input$0.59$0.351$0.06 (best)
Output$0.745$0.555$0.24 (best)
Cached input——$0.015
Blended (3:1)$0.629$0.402$0.105 (best)
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 40 providersMedian of 9 providersOfficial Amazon Bedrock API
Limits
Context window128,000 tokens128,000 tokens300,000 tokens (best)
Max output4,096 tokens16,384 tokens (best)10,000 tokens
Inputs and features
TextYesYesYes
ImagesNoYesYes
PDFsNoNoYes
AudioNoNoNo
VideoNoNoYes
ReasoningNoNoNo
Tool callingYesYesYes
Structured outputNoNoNo
Availability
WeightsOpenOpenProprietary
API model ID——amazon.nova-lite-v1:0
API providers43 (best)98
ReleasedDec 6, 2024Mar 17, 2025Dec 3, 2024
Knowledge cutoffDec 2023Jun 2024Oct 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.

  • Llama-3.3-70B-Instruct$7.39
  • Mistral Small 3.1 24B$4.62
  • Nova Lite$1.08
04 — Questions

Which should you choose?

Which is better: Llama-3.3-70B-Instruct, Mistral Small 3.1 24B or Nova Lite?

There is not enough shared capability evidence to name an overall winner. The scores (Nova Lite 60/100, Mistral Small 3.1 24B 43/100, Llama-3.3-70B-Instruct 28/100) compare the available specifications only. The score weighs inputs & features 60%, context window 40%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

Which is cheaper, Llama-3.3-70B-Instruct, Mistral Small 3.1 24B or Nova Lite?

Nova Lite is cheaper at $0.06 input / $0.24 output per million tokens (official Amazon Bedrock API price). Mistral Small 3.1 24B costs $0.351 input / $0.555 output per million tokens (median across 9 API providers); Llama-3.3-70B-Instruct costs $0.59 input / $0.745 output per million tokens (median across 40 API providers). At a typical mix of three input tokens to one output token, that is $0.105 per million tokens for Nova Lite versus $0.402 for Mistral Small 3.1 24B (3.8× as much) and $0.629 for Llama-3.3-70B-Instruct (6× as much).

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Llama-3.3-70B-Instruct has an ECI of 126.2, Mistral Small 3.1 24B has not been scored yet and Nova Lite has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Llama-3.3-70B-Instruct, Mistral Small 3.1 24B and Nova Lite yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.

Which has the bigger context window?

Nova Lite has the largest context window at 300,000 tokens, against 128,000 for Llama-3.3-70B-Instruct and 128,000 for Mistral Small 3.1 24B. Maximum output per response: Llama-3.3-70B-Instruct up to 4,096, Mistral Small 3.1 24B up to 16,384, Nova Lite up to 10,000 tokens.

Which can read images, PDFs, audio or video?

Llama-3.3-70B-Instruct accepts text; Mistral Small 3.1 24B accepts text and images; Nova Lite accepts text, images, PDFs and video. Nova Lite handles the widest range of inputs.

Are any of these open source?

Llama-3.3-70B-Instruct and Mistral Small 3.1 24B publishes its weights and can be self-hosted; Nova Lite is proprietary.

Which is newer?

Mistral Small 3.1 24B is the newest, released Mar 17, 2025. Llama-3.3-70B-Instruct came out Dec 6, 2024; Nova Lite came out Dec 3, 2024. Knowledge cutoff: Llama-3.3-70B-Instruct Dec 2023, Mistral Small 3.1 24B Jun 2024, Nova Lite Oct 2024.

How do you decide the winner?

This rating uses inputs & features (60%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (40%, 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 18, 2026.