ModelSpec is an analysis of alternatives for AI models: it decides which models fit a job from your requirements, sourced benchmarks and real cost, and shows its work.

Maths and reasoning, balanced

Find an active text model strong on maths and multi-step reasoning without ignoring task cost.

Worked example as of 2026-09-23, from the public data image.

Maths and reasoning together, with cost as a third term.

Requirements

model.class = text-generator: The answer must be generated as text.
model.lifecycle = active: An active model is still supported by its lab or provider.

Preferences

maths: Prefer stronger evidence on objectively checked maths.
reasoning: Prefer stronger multi-step reasoning.
-offering.cost_per_task: Among strong reasoners, prefer the cheaper task.

Applied template spec

spec_version

where

model.class = text-generator
model.lifecycle = active

optimize

weights

maths: 0.4
reasoning: 0.3
-offering.cost_per_task: 0.3

Worked answer

Decision status: partial.

No recommendation: the public image has no answer with enough evidence.

This is a partial answer. Unknown facts or missing objective evidence can leave eligible candidates unranked.

probability bands: a model joins the leader when P(its score >= the leader's) >= 0.25 under the score posteriors; a weighted capability with an 80% interval wider than 2.8 is not enough evidence

Score and score interval use the objective's dimensionless normalised units. Cost per task is for the engine-selected offering and the applied spec's task size. When task size is omitted, the engine uses its documented default. P(best) is the engine probability. Unavailable means the engine did not return a value. Thin evidence is undetermined and is not a recommendation.

Thin evidence: undetermined

Model	Score	Score interval	P(best)	Cost per task, USD	Capability estimates and intervals
`openai/gpt-5-4`	0.750341	[0.369581, 1.131102]	0.12109375	0.16	maths: 0.925643173; interval [0.204892258, 1.646394088]; benchmarks 5; direct benchmarks 2 reasoning: 0.230107427; interval [-1.318205455, 1.77842031]; benchmarks 3; direct benchmarks 0
`google/gemini-3-7-flash`	0.718987	[0.276332, 1.161642]	0.08984375	0.045	maths: -0.018394566; interval [-0.971933015, 0.935143883]; benchmarks 3; direct benchmarks 1 reasoning: 0.463255629; interval [-1.246777358, 2.173288616]; benchmarks 2; direct benchmarks 0
`openai/gpt-5-6-sol`	0.693345	[0.24477, 1.14192]	0.078125	0.24	maths: 0.917782921; interval [-0.058069444, 1.893635285]; benchmarks 3; direct benchmarks 1 reasoning: 0.148460441; interval [-1.577008032, 1.873928913]; benchmarks 2; direct benchmarks 0
`anthropic/claude-opus-5`	0.69266	[0.246538, 1.138782]	0.09765625	0.3	maths: 0.740109404; interval [-0.226442817, 1.706661626]; benchmarks 3; direct benchmarks 1 reasoning: 0.45386262; interval [-1.265252052, 2.172977292]; benchmarks 2; direct benchmarks 0
`google/gemini-3-8-flash`	0.676751	[0.307733, 1.045768]	0.07421875	0.045	maths: -0.457512924; interval [-1.157546025, 0.242520177]; benchmarks 5; direct benchmarks 2 reasoning: 0.570955383; interval [-0.928430754, 2.070341519]; benchmarks 3; direct benchmarks 0
`google/gemini-3-1-pro-preview`	0.650517	[0.277124, 1.02391]	0.0625	0.128	maths: 0.10766147; interval [-0.613089445, 0.828412385]; benchmarks 5; direct benchmarks 2 reasoning: 0.22241102; interval [-1.285090676, 1.729912715]; benchmarks 3; direct benchmarks 0
`openai/gpt-6-astra`	0.63018	[0.258443, 1.001917]	0.046875	0.6	maths: 1.561674136; interval [0.843459177, 2.279889095]; benchmarks 4; direct benchmarks 2 reasoning: 0.323374241; interval [-1.176927713, 1.823676194]; benchmarks 3; direct benchmarks 0
`meta/muse-spark-1-3`	0.6155	[0.123931, 1.107068]	0.08984375	0.067	maths: -0.351147097; interval [-1.279192008, 0.576897813]; benchmarks 3; direct benchmarks 1 reasoning: 0.211605235; interval [-1.789205691, 2.212416162]; benchmarks 1; direct benchmarks 0
`zhipu/glm-5-3`	0.582196	[0.141293, 1.023099]	0.05859375	0.0736	maths: -0.156897477; interval [-1.103624192, 0.789829238]; benchmarks 3; direct benchmarks 1 reasoning: -0.105195063; interval [-1.810826758, 1.600436632]; benchmarks 2; direct benchmarks 0
`meta/muse-spark-1-1`	0.574653	[-0.067599, 1.216905]	0.0859375	0.067	maths: 0.07173741; interval [-1.929073517, 2.072548336]; benchmarks 1; direct benchmarks 0 reasoning: -0.343673105; interval [-2.344484032, 1.657137821]; benchmarks 1; direct benchmarks 0
`google/gemini-3-5-flash`	0.559137	[0.1401, 0.978174]	0.02734375	0.096	maths: -0.555304745; interval [-1.295683765, 0.185074274]; benchmarks 4; direct benchmarks 2 reasoning: 0.140878753; interval [-1.604182337, 1.885939842]; benchmarks 2; direct benchmarks 0
`anthropic/claude-fable-5-1`	0.530454	[0.128877, 0.932032]	0.0546875	0.6	maths: 0.354621934; interval [-0.329488136, 1.038732003]; benchmarks 6; direct benchmarks 2 reasoning: 0.710115622; interval [-0.98201435, 2.402245595]; benchmarks 2; direct benchmarks 0
`anthropic/claude-opus-4-6`	0.529791	[0.156398, 0.903184]	0.0234375	0.3	maths: -0.089733789; interval [-0.810484704, 0.631017127]; benchmarks 5; direct benchmarks 2 reasoning: 0.197361213; interval [-1.310140483, 1.704862909]; benchmarks 3; direct benchmarks 0
`anthropic/claude-fable-5`	0.477755	[0.028725, 0.926786]	0.02734375	0.6	maths: 0.841368906; interval [-0.153084238, 1.83582205]; benchmarks 3; direct benchmarks 1 reasoning: 0.039498036; interval [-1.674020333, 1.753016406]; benchmarks 2; direct benchmarks 0
`anthropic/claude-opus-4-7`	0.459648	[0.086222, 0.833075]	0.0078125	0.3	maths: -0.184419351; interval [-0.905407051, 0.536568349]; benchmarks 5; direct benchmarks 2 reasoning: -0.117498278; interval [-1.624999974, 1.390003417]; benchmarks 3; direct benchmarks 0
`deepseek/deepseek-v4-pro`	0.430597	[0.01363, 0.847565]	0.01953125	0.06864	maths: -0.735823404; interval [-1.474402989, 0.00275618]; benchmarks 4; direct benchmarks 2 reasoning: -0.508628162; interval [-2.24362481, 1.226368486]; benchmarks 2; direct benchmarks 0
`zhipu/glm-5-2`	0.421507	[0.005205, 0.837809]	0.0078125	0.0736	maths: -1.082359497; interval [-1.820422979, -0.344296015]; benchmarks 4; direct benchmarks 2 reasoning: -0.275228659; interval [-2.006940496, 1.456483179]; benchmarks 2; direct benchmarks 0
`google/gemini-2-5-flash`	0.386419	[-0.183581, 0.956418]	0.02734375	0.022	maths: -0.651775501; interval [-2.147773284, 0.844222281]; benchmarks 3; direct benchmarks 0 reasoning: -0.95162632; interval [-2.944986922, 1.041734282]; benchmarks 2; direct benchmarks 0

Unknown requirements

Some candidates may qualify. Missing facts:

maths
offering.cost_per_task
reasoning

Engine warnings

not_separable

Omitted result rows: models 6; offerings 11. The bands above include all ranked models.

Decision dec_1c52f84dee09a6e6c91dbcbf on public snapshot snap_05a963d2477ac2fb.

What this template teaches

Two capability Prefers can share the ranking with cost; each keeps its own share.

The live board uses current data. Its answer can change as evidence changes.

Apply this template on Decide

Human lookup is free on Decide. CLI, API, and MCP access requires a key. Access and pricing.

Use-case examples · How we decide