Cheaper alternatives to Olmo 3 32B Think

Allen AI LMArena 1298.5 $0.237/M on a balanced workload prices as of 2026-08-28

11 models are both better and cheaper, giving up nothing.

Of the 133 models carrying both an independent score and a published price at the same delivery mode, 22 score at least as high as Olmo 3 32B Think and cost no more under at least one workload. 11 match or beat it on every capability we hold — context, maximum output, input modes, tool use and extended reasoning. 11 do not, and every row names what it drops.

Dominance is not a property of a model. It is a property of a model, a capability lens and a workload, and all three are named on every row. Scores are LMArena Elo, used under CC BY 4.0 from the official dataset; prices are blended per million tokens. A higher score is not a drop-in replacement.

This page is built from that comparison and nothing else. When no model is both better and cheaper than Olmo 3 32B Think, it is not generated.

Where it holds

Workload	Olmo 3 32B Think $/M	Better and cheaper	With a trade
Balanced	$0.237	8	9
Summarise	$0.168	9	10
Chat	$0.290	9	9
Code gen	$0.360	10	8
Agentic	$0.203	10	10

Workload mixes are defined on the methodology page. Cached input is priced at the cached rate, and reasoning tokens at the worse of the reasoning and output rates.

What you would be replacing

Intelligence	1298.5
Context window	66K
Max output	59K
Input modes	text
Tool use	no
Extended reasoning	yes

A replacement has to clear every line above, not just the score. Full record for Olmo 3 32B Think.

Better and cheaper, nothing given up 11

Each of these matches or beats Olmo 3 32B Think on context, maximum output, input modes, tool use and reasoning, scores at least as high, and costs no more.

#	Model	Intelligence	$/M	Holds under
1	Hy3 Tencent	1441.2 +142.7	$0.144	Balanced −39% Summarise −54% Chat −44% Code gen −37% Agentic −59%
2	DeepSeek V4 Flash 0423 DeepSeek	1431.6 +133.1	$0.101	Balanced −57% Summarise −60% Chat −68% Code gen −65% Agentic −73%
3	MiMo-V2.5 Xiaomi	1427.3 +128.8	$0.175	Balanced −26% Summarise −36% Chat −47% Code gen −41% Agentic −61%
4	Step 3.5 Flash StepFun	1403.8 +105.3	$0.150	Balanced −37% Summarise −34% Chat −38% Code gen −39% Agentic −36%
5	Qwen3.5-Flash Qwen	1397.6 +99.1	$0.114	Balanced −52% Summarise −55% Chat −51% Code gen −49% Agentic −53%
6	Solar Pro 4 Upstage	1376.2 +77.7	$0.052	Balanced −78% Summarise −83% Chat −80% Code gen −77% Agentic −86%
7	gpt-oss-120b OpenAI	1365.6 +67.1	$0.070	Balanced −70% Summarise −74% Chat −69% Code gen −68% Agentic −72%
8	GPT-5 Nano OpenAI	1320.3 +21.8	$0.138	Balanced −42% Summarise −67% Chat −39% Code gen −29% Agentic −63%
9	DeepSeek V3.2 DeepSeek	1424.6 +126.1	$0.269 chat	Chat −7% Code gen −11% Agentic −1%
10	DeepSeek V3.2 Exp DeepSeek	1424.4 +125.9	$0.354 code gen	Code gen −2%
11	GLM 4.5 Air Z.ai	1382.8 +84.3	$0.136 summarise	Summarise −19% Agentic −13%

Cheaper and higher-scoring, but you give something up 11

These score at least as high and cost no more on the two plotted axes, and lose something that is not on them. Read the last column before switching.

Model	Intelligence	$/M	Holds under	What you give up
Gemma 4 31B Google	1441.7 +143.2	$0.152	Balanced −36% Summarise −46% Chat −39% Code gen −34% Agentic −49%	59K → 16K max output
Gemma 4 26B A4B Google	1434.6 +136.1	$0.138	Balanced −42% Summarise −50% Chat −39% Code gen −36% Agentic −45%	59K → 16K max output
Qwen3 235B A22B Instruct 2507 Qwen	1419.3 +120.8	$0.205	Balanced −14% Summarise −33% Chat −6% Agentic −21%	59K → 16K max output no extended reasoning
Qwen3 30B A3B Instruct 2507 Qwen	1384.3 +85.8	$0.084	Balanced −64% Summarise −67% Chat −63% Code gen −62% Agentic −65%	59K → 32K max output no extended reasoning
Gemma 3 27B Google	1358.3 +59.8	$0.172	Balanced −27% Summarise −48% Chat −26% Code gen −17% Agentic −45%	no extended reasoning
GLM 4.7 Flash Z.ai	1352.9 +54.4	$0.145	Balanced −39% Summarise −63% Chat −38% Code gen −28% Agentic −60%	59K → 16K max output
Qwen3 32B Qwen	1340.1 +41.6	$0.130	Balanced −45% Summarise −46% Chat −45% Code gen −44% Agentic −46%	59K → 16K max output
Gemma 3 12B Google	1334.2 +35.7	$0.075	Balanced −68% Summarise −67% Chat −69% Code gen −69% Agentic −68%	59K → 16K max output no extended reasoning
Qwen3 30B A3B Qwen	1316.9 +18.4	$0.215	Balanced −9% Summarise −17% Chat −6% Code gen −3% Agentic −13%	59K → 16K max output
Qwen3 Next 80B A3B Instruct Qwen	1418.6 +120.1	$0.141 summarise	Summarise −16%	no extended reasoning
Mercury 2 Inception	1357.8 +59.3	$0.191 agentic	Agentic −6%	59K → 50K max output

What this compares, and what it leaves out

Quality is LMArena Elo, used under CC BY 4.0 from the official dataset. The 95% confidence interval on a difference between two scores is about ±10.68 points, so a gap smaller than that is marked tie rather than an improvement — see significance bands.
188 further models at this delivery mode carry a price but no independent score. They are absent from the comparison in both directions — unrated is not a zero, and an unmeasured model is neither an alternative nor a worse buy.
Retired models are never offered as an alternative, and a model only competes against its own delivery mode: batch trades latency for price, so it is not a like-for-like swap.
Nothing here measures latency, throughput, rate limits or how a model behaves on your prompts. Two models with the same index score are not interchangeable.
The models that nothing beats on both axes are on the value frontier, and every other model something cheaper beats is listed here.