NVIDIA Nemotron 3 Ultra 550B A55B vs Qwen3 235B A22B Instruct 2507

Close to a tie overall. Qwen3 235B A22B Instruct 2507 costs 78% less for the same work, and both read a 256K-token context.

Model A

NVIDIA Nemotron 3 Ultra 550B A55B

by NVIDIA · #26 of 67 overall

Coming soon · price at launch

Context: 256K
Input / 1M: $0.50
Output / 1M: $2.20

Model B

Qwen3 235B A22B Instruct 2507

by Qwen · #28 of 67 overall

Coming soon · price at launch

Context: 256K
Input / 1M: $0.090
Output / 1M: $0.55

Better overall

About even

Grade 72.8 vs 71.0

Cheaper

Qwen3 235B A22B Instruct 2507

78% less, $0.20 vs $0.93 blended

Larger context

Same

256K vs 256K tokens

Head to head

Public benchmarks. Longer is better; each value links to its source.

LMArena TextElo-style rating · LMArena (CC BY 4.0) · Oct 15, 2026

NVIDIA Nemotron 3 Ultra 550B A55B1,426

Qwen3 235B A22B Instruct 25071,422

LMArena CodingElo-style rating · LMArena (CC BY 4.0) · Oct 15, 2026

NVIDIA Nemotron 3 Ultra 550B A55B1,474

Qwen3 235B A22B Instruct 25071,472

Aider Polyglot% of exercises solved · Aider (aider.chat) · Jul 21, 2025

NVIDIA Nemotron 3 Ultra 550B A55B—

Qwen3 235B A22B Instruct 250757.3%^v

SWE-bench Verified% of issues resolved · SWE-bench · Jun 4, 2026 · reference

NVIDIA Nemotron 3 Ultra 550B A55B70.7%^v

Qwen3 235B A22B Instruct 2507—

Terminal-Bench 2.1% of tasks passed · Terminal-Bench · Jun 4, 2026 · reference

NVIDIA Nemotron 3 Ultra 550B A55B56.4%^v

Qwen3 235B A22B Instruct 2507—

LiveCodeBenchpass@1, % · LiveCodeBench · Jul 21, 2025 · reference

NVIDIA Nemotron 3 Ultra 550B A55B—

Qwen3 235B A22B Instruct 250751.8%^v

NVIDIA Nemotron 3 Ultra 550B A55BQwen3 235B A22B Instruct 2507v = reported by the model’s maker. — = no public score yet. Arena bars span the range across every model we grade.

What it costs you

Drag to match your monthly usage.

Input tokens a month200MOutput tokens a month40M

NVIDIA Nemotron 3 Ultra 550B A55B$188 /month

Qwen3 235B A22B Instruct 2507$40.00 /month

Qwen3 235B A22B Instruct 2507 saves you $148 a month at this usage, $1,776 a year.

At a glance

Attribute	NVIDIA Nemotron 3 Ultra 550B A55B	Qwen3 235B A22B Instruct 2507
Made by	NVIDIA	Qwen
Availability	Coming soon	Coming soon
Overall grade	Excellent · 72.8	Excellent · 71.0
Overall rank	#26 of 67	#28 of 67
Input price	$0.50 per 1M	$0.090 per 1M
Output price	$2.20 per 1M	$0.55 per 1M
Context window	262,144 tokens	262,144 tokens
Reasoning levels	No	Yes
Reads images	No	No
Model ID	nvidia-nemotron-3-ultra-550b-a55b	qwen3-235b-a22b-instruct-2507

Try both with one key

Same endpoint, same key, both hosted in the USA. Change one word to switch models, in Claude Code, Cursor, Cline or your own code.

Get your API key

# Ask either model the same thing
curl https://fullsendai.com/v1/chat/completions \
  -H "Authorization: Bearer $FULLSEND_API_KEY" \
  -d '{"model": "nvidia-nemotron-3-ultra-550b-a55b", "messages": [...]}'

# …then swap one word
  -d '{"model": "qwen3-235b-a22b-instruct-2507", "messages": [...]}'

Questions

Is NVIDIA Nemotron 3 Ultra 550B A55B or Qwen3 235B A22B Instruct 2507 better?

They grade within two points of each other (72.8 and 71.0), so pick on price and the benchmarks that match your work.

Which is cheaper, NVIDIA Nemotron 3 Ultra 550B A55B or Qwen3 235B A22B Instruct 2507?

Qwen3 235B A22B Instruct 2507: $0.090 in and $0.55 out per million tokens, against $0.50 and $2.20 through FullsendAI.

Which has the larger context window?

Neither: both take 262,144 tokens.

Can I use both with one API key?

Yes, once they launch: both are coming soon, at the prices shown, and both will use the same FullsendAI key.

Model rankings All benchmarks Compare other models

Scores: LMArena (CC BY 4.0), Aider (aider.chat), SWE-bench, Terminal-Bench, LiveCodeBench, and model makers’ published results where marked v. Prices are FullsendAI’s, per million tokens. Data as of Oct 28, 2026.