Skip to main content

Model comparison

Gemma 4 31B vs Inkling

Data verified

Head-to-head evidence from 4 shared benchmark results across 2 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

61/100
Margin
8.0pts
winning →
Thinking Machines Lab
69/100
2 category wins1 category wins

Verified leaderboard positions: Gemma 4 31B unranked; Inkling #17

BenchAlign evidence: Gemma 4 31B supported; Inkling not scored. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Gemma 4 31B and Inkling share 4 comparable benchmark results. 3 of 8 categories are comparable. 26 results are unique to Gemma 4 31B; 12 to Inkling.

Updated July 15, 2026
Shared results
4
Gemma 4 31B only
26
Inkling only
12
Comparable categories
3 / 8

Pick Inkling if you want the stronger benchmark profile. Gemma 4 31B only becomes the better choice if knowledge is the priority or you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 4 shared benchmark results across 2 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Inkling is clearly ahead on the provisional aggregate, 69 to 62. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Inkling's sharpest advantage is in coding, where it averages 68.6 against 41.6. The single biggest benchmark swing on the page is HLE, 26.5% to 46%. Gemma 4 31B does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.

Inkling is also the more expensive model on tokens at $1.87 input / $4.68 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Gemma 4 31B. That is roughly Infinityx on output cost alone. Gemma 4 31B is the reasoning model in the pair, while Inkling is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Inkling gives you the larger context window at 1M, compared with 256K for Gemma 4 31B.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Gemma 4 31B and Inkling
CategoryGemma 4 31BΔInkling
CodingGemma 4 31B41.6Margin 27.0Inkling68.6
KnowledgeGemma 4 31B53.3Margin 1.6Inkling51.7
MultimodalGemma 4 31B76.9Margin 0.4Inkling76.5
AgenticGemma 4 31BNot measuredMarginNo overlapInkling69.4
MathGemma 4 31BNot measuredMarginNo overlapInkling97.1
Inst. FollowingGemma 4 31BNot measuredMarginNo overlapInkling79.8

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Gemma 4 31BB · Inkling
  1. HLE

    Knowledge
    Source ↗
    A 26.5%B 46%
    Winner: InklingΔ 19.5
    HLE: Gemma 4 31B scored 26.5%; Inkling scored 46%. Inkling wins this benchmark.
  2. GPQA

    Knowledge
    Source ↗
    A 84.3%B 87.9%
    Winner: InklingΔ 3.6
    GPQA: Gemma 4 31B scored 84.3%; Inkling scored 87.9%. Inkling wins this benchmark.
  3. MMMU-Pro

    Multimodal
    Source ↗
    A 76.9%B 73.5%
    Winner: Gemma 4 31BΔ 3.4
    MMMU-Pro: Gemma 4 31B scored 76.9%; Inkling scored 73.5%. Gemma 4 31B wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricGemma 4 31BInklingComparison
Input / output priceUSD per 1M tokensGemma 4 31B$0 input / $0 outputInkling$1.87 input / $4.68 outputGemma 4 31B has the lower combined listed price.
Generation speedtokens per secondGemma 4 31BNot availableInklingNot availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenGemma 4 31BNot availableInklingNot availableA complete latency comparison is not available.
Context windowmaximum listed tokensGemma 4 31B256KInkling1MInkling lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkGemma 4 31BInklingResult
AA Agentic IndexSource 14.4%Not comparable
τ²-bench resultsSource 59.9%Not comparable
GDPval-AASource 15.2%Not comparable
GDPval-AASource 804Not comparable
Gert LabsSource 35.26%Not comparable
AA EnterpriseOps-GymSource 28.3%Not comparable
AA Harvey LABSource 0.0%Not comparable
AA ITBenchSource 37.3%Not comparable
AA Tau3 BankingSource 15.1%Not comparable
Terminal-Bench 2.0Source 63.8%Not comparable
BrowseCompSource 77.1%Not comparable
MCP AtlasSource 74.1%Not comparable
Design Arena Agentic Web DevSource 1258Not comparable
CodingInkling wins
BenchmarkGemma 4 31BInklingResult
SWE-RebenchSource 41.6%Not comparable
React Native EvalsSource 75.2%Not comparable
AA Coding IndexSource 43.4%Not comparable
Terminal-Bench HardSource 36.4%Not comparable
AA-SciCodeSource 43.4%Not comparable
SWE-bench VerifiedSource 77.6%Not comparable
SWE-bench ProSource 54.3%Not comparable
Terminal-Bench 2.0Source 63.8%Not comparable
Reasoning
BenchmarkGemma 4 31BInklingResult
AA-LCRSource 62.0%Not comparable
CritPtSource 1.4%Not comparable
KnowledgeGemma 4 31B wins
BenchmarkGemma 4 31BInklingResult
GPQASource 84.3%87.9%Inkling leads
MMLU-ProSource 85.2%Not comparable
HLESource 26.5%46%Inkling leads
HLE w/o toolsSource 19.5%30%Inkling leads
Artificial Analysis Intelligence IndexSource 29.4%Not comparable
AA-GPQA DiamondSource 85.7%Not comparable
AA-HLESource 22.7%Not comparable
AA-Omniscience IndexSource -45.4%Not comparable
AA-Omniscience AccuracySource 19.9%Not comparable
AA-Omniscience Hallucination RateSource 81.6%Not comparable
AA Openness IndexSource 38.9%Not comparable
GPQA-DSource 87.9%Not comparable
Math
BenchmarkGemma 4 31BInklingResult
AIME26Source 97.1%Not comparable
MultimodalGemma 4 31B wins
BenchmarkGemma 4 31BInklingResult
MMMU-ProSource 76.9%73.5%Gemma 4 31B leads
AA-MMMU-ProSource 73.4%Not comparable
CharXivSource 82%Not comparable
CharXiv w/o toolsSource 78.1%Not comparable
Inst. Following
BenchmarkGemma 4 31BInklingResult
AA-IFBenchSource 75.6%Not comparable
IFBenchSource 79.8%Not comparable
Frequently Asked Questions (4)

Which is better, Gemma 4 31B or Inkling?

Inkling is ahead on BenchLM's provisional leaderboard, 69 to 62. The biggest single separator in this matchup is HLE, where the scores are 26.5% and 46%.

Which is better for knowledge tasks, Gemma 4 31B or Inkling?

Gemma 4 31B has the edge for knowledge tasks in this comparison, averaging 53.3 versus 51.7. Inside this category, HLE is the benchmark that creates the most daylight between them.

Which is better for coding, Gemma 4 31B or Inkling?

Inkling has the edge for coding in this comparison, averaging 68.6 versus 41.6. Gemma 4 31B stays close enough that the answer can still flip depending on your workload.

Which is better for multimodal and grounded tasks, Gemma 4 31B or Inkling?

Gemma 4 31B has the edge for multimodal and grounded tasks in this comparison, averaging 76.9 versus 76.5. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

Gemma 4 31B
API / mo$0
Self-host / mo$429
Break-even
Inkling
API / mo$4,913
Self-host / moNot listed
Break-even
Proprietary model — self-hosting not applicable.
Model the full break-even

Related Comparisons

Last updated: July 15, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.