ZMIME
Comparison · 3 models · Updated Oct 4, 2026

Llama 3.1 Nemotron 70B Instruct vs Llama 3.1 Nemotron Ultra 253B vs Qwen3-VL 30B-A3B

Llama 3.1 Nemotron Ultra 253B comes out ahead, 65 to 59 and 45 on our weighted score, and it is the cheaper option too.

  1. NVIDIA

    Llama 3.1 Nemotron 70B Instruct

    Released Apr 15, 2025

    45/100
    • ECI—
    • Price$0.478 / $0.504
    • Context128K
    Full specs Remove Llama 3.1 Nemotron 70B Instruct
  2. Our pick

    NVIDIA

    Llama 3.1 Nemotron Ultra 253B

    Released Apr 7, 2025

    65/100
    • ECI—
    • PriceFree / Free
    • Context128K
    Full specs Remove Llama 3.1 Nemotron Ultra 253B

  3. Alibaba (Qwen)

    Qwen3-VL 30B-A3B

    Released Apr 2025

    59/100
    • ECI—
    • Price$0.20 / $0.80
    • Context131K
    Full specs Remove Qwen3-VL 30B-A3B
01 — Verdict

Llama 3.1 Nemotron Ultra 253B is our pick

Llama 3.1 Nemotron Ultra 253B is the better all-round choice, scoring 65/100 against Qwen3-VL 30B-A3B (59) and Llama 3.1 Nemotron 70B Instruct (45). It leads on price. Qwen3-VL 30B-A3B wins on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. 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 priceLlama 3.1 Nemotron Ultra 253BLlama 3.1 Nemotron Ultra 253B Free · Qwen3-VL 30B-A3B $0.35 · Llama 3.1 Nemotron 70B Instruct $0.485 per 1M tokens (3:1 blend)
  • Longest contextQwen3-VL 30B-A3BQwen3-VL 30B-A3B 131,072 · Llama 3.1 Nemotron 70B Instruct 128,000 · Llama 3.1 Nemotron Ultra 253B 128,000 tokens
  • Widest inputsQwen3-VL 30B-A3BLlama 3.1 Nemotron 70B Instruct: Text · Llama 3.1 Nemotron Ultra 253B: Text · Qwen3-VL 30B-A3B: Text, Images
  • Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
How the score is built
MeasureWeightLlama 3.1 Nemotron 70B InstructLlama 3.1 Nemotron Ultra 253BQwen3-VL 30B-A3B
Price50%6510072
Inputs & features30%253560
Context window20%242424
Overall100%45/10065/10059/100

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 3.1 Nemotron 70B Instruct vs Llama 3.1 Nemotron Ultra 253B vs Qwen3-VL 30B-A3B specifications side by side
SpecificationLlama 3.1 Nemotron 70B InstructNVIDIALlama 3.1 Nemotron Ultra 253BNVIDIAQwen3-VL 30B-A3BAlibaba (Qwen)
Capability
Capabilities Index (ECI)———
ECI rank———
Price per million tokens
Input$0.478Free (best)$0.20
Output$0.504Free (best)$0.80
Cached input———
Blended (3:1)$0.485Free (best)$0.35
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 2 providersOfficial Nvidia APIOfficial Alibaba API
Limits
Context window128,000 tokens128,000 tokens131,072 tokens (best)
Max output8,192 tokens8,192 tokens32,768 tokens (best)
Inputs and features
TextYesYesYes
ImagesNoNoYes
PDFsNoNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningNoYesYes
Tool callingYesYesYes
Structured outputNoNoNo
Availability
WeightsOpenOpenOpen
API model IDnvidia/llama-3.1-nemotron-70b-instructnvidia/llama-3.1-nemotron-ultra-253b-v1qwen3-vl-30b-a3b
API providers3 (best)11
ReleasedApr 15, 2025Apr 7, 2025Apr 2025
Knowledge cutoff——Apr 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.

  • Llama 3.1 Nemotron 70B Instruct$5.79
  • Llama 3.1 Nemotron Ultra 253BFree
  • Qwen3-VL 30B-A3B$3.60
04 — Questions

Which should you choose?

Which is better: Llama 3.1 Nemotron 70B Instruct, Llama 3.1 Nemotron Ultra 253B or Qwen3-VL 30B-A3B?

Llama 3.1 Nemotron Ultra 253B is the better all-round choice, scoring 65/100 against Qwen3-VL 30B-A3B (59) and Llama 3.1 Nemotron 70B Instruct (45). It leads on price. Qwen3-VL 30B-A3B wins on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

Which is cheaper, Llama 3.1 Nemotron 70B Instruct, Llama 3.1 Nemotron Ultra 253B or Qwen3-VL 30B-A3B?

Llama 3.1 Nemotron Ultra 253B is cheaper at Free input / Free output per million tokens (official Nvidia API price). Qwen3-VL 30B-A3B costs $0.20 input / $0.80 output per million tokens (official Alibaba API price); Llama 3.1 Nemotron 70B Instruct costs $0.478 input / $0.504 output per million tokens (median across 2 API providers; free on Nvidia). Llama 3.1 Nemotron Ultra 253B is listed as free.

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Llama 3.1 Nemotron 70B Instruct has not been scored yet, Llama 3.1 Nemotron Ultra 253B has not been scored yet and Qwen3-VL 30B-A3B has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Llama 3.1 Nemotron 70B Instruct, Llama 3.1 Nemotron Ultra 253B and Qwen3-VL 30B-A3B 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?

Qwen3-VL 30B-A3B has the largest context window at 131,072 tokens, against 128,000 for Llama 3.1 Nemotron 70B Instruct and 128,000 for Llama 3.1 Nemotron Ultra 253B. Maximum output per response: Llama 3.1 Nemotron 70B Instruct up to 8,192, Llama 3.1 Nemotron Ultra 253B up to 8,192, Qwen3-VL 30B-A3B up to 32,768 tokens.

Which can read images, PDFs, audio or video?

Llama 3.1 Nemotron 70B Instruct accepts text; Llama 3.1 Nemotron Ultra 253B accepts text; Qwen3-VL 30B-A3B accepts text and images. Qwen3-VL 30B-A3B handles the widest range of inputs.

Are any of these open source?

Yes, all three publish their weights, so you can self-host them.

Which is newer?

Llama 3.1 Nemotron 70B Instruct is the newest, released Apr 15, 2025. Llama 3.1 Nemotron Ultra 253B came out Apr 7, 2025; Qwen3-VL 30B-A3B came out Apr 2025. Knowledge cutoff: Qwen3-VL 30B-A3B 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 Oct 4, 2026.

05 — Keep comparing

Related comparisons