Gemini 3.7 Flash: Benchmarks, Pricing & Verdict

Gemini 3.7 Flash is Google's latest workhorse model, positioned explicitly for coding and agentic workflows rather than as a frontier reasoning flagship - that role still belongs to the yet-unreleased Gemini 3.5 Pro tier this generation. Announced by Tulsee Doshi, Google's Senior Director of Product Management for Gemini, the model builds on Gemini 3.6 Flash's foundation with what Google describes as algorithmic improvements to core reasoning rather than a full retrain, which is consistent with both the three-week gap since the last release and the model card's decision to defer architectural detail to the 3.6 Flash documentation.

On Google's own published evaluation table - which includes Gemini 3.6 Flash, Claude Sonnet 5, GPT-5.6 Terra and Muse Spark 1.2 as comparison points - Gemini 3.7 Flash improves on its predecessor across essentially every listed benchmark, sometimes by a wide margin (DeepSWE v1.1 jumps from 48.6% to 65.3%; AutomationBench nearly doubles from 17.0% to 30.4%). Against outside competitors the picture is more mixed: it leads on several agentic and document-comprehension benchmarks but trails GPT-5.6 Terra on the composite Artificial Analysis Intelligence Index and on two coding-specific benchmarks. That nuance is worth holding onto, because Google's own marketing compresses it into a single headline claim.

• Best for: coding agents, terminal-based agentic workflows, and long-document or long-context enterprise tasks where Gemini 3.7 Flash's specific benchmark strengths (Code Arena, GDM-MRCR long-context recall, Harvey legal-document work) line up with the job.
• Headline price: $0.75 input / $3.75 output per million tokens (about £0.59 / £2.95) through 31 December 2026 - unchanged from 3.6 Flash's launch pricing, not a price cut.
• Headline benchmark move: DeepSWE v1.1 long-horizon software engineering up from 48.6% to 65.3% versus its own predecessor; still 4.3 points behind GPT-5.6 Terra's 69.6% on the same benchmark.
• Main caveat: Google's claim of beating rivals "across nine benchmarks" is accurate but partial - GPT-5.6 Terra leads the Artificial Analysis Intelligence Index, DeepSWE, both Terminal-Bench versions and OSWorld-2.0 on the same published table.

Gemini 3.6 Flash launched alongside Gemini 3.5 Flash-Lite and a specialised Gemini 3.5 Flash Cyber variant. Three weeks later, Gemini 3.7 Flash replaces the mid-tier Flash model in that lineup while the flagship Gemini 3.5 Pro tier - and the much-discussed Gemini 4 generation covered separately on this site - remain unreleased at time of writing. That means Gemini 3.7 Flash is, unusually, shipping ahead of the larger reasoning model in its own generation, which is part of why Google frames it specifically as "our most intelligent workhorse model" rather than as a flagship replacement.

Google attributed the accelerated schedule to developer feedback and algorithmic innovations in the model's reasoning foundation, language that suggests targeted post-training improvements on the existing 3.6 Flash base rather than a ground-up retrain - consistent with the model card explicitly deferring architecture documentation to the 3.6 Flash card and reporting the same knowledge cutoff (March 2026, with some domains only current to January 2025). For a lab racing to ship, reusing a base model and iterating on post-training is the more plausible explanation than three weeks being enough time to pretrain a new model from scratch.

Gemini 3.7 Flash supports a 1-million-token input context window and generates up to 64,000 tokens of output per response, in line with the rest of the Gemini 3 family. It ships with customisable "thinking" configuration, letting developers trade latency for reasoning depth per request rather than being locked into a single reasoning budget.

Availability at launch spans the Gemini API and Google AI Studio, Android Studio for mobile developers, Google Antigravity - Google's agent-first coding environment - the Gemini Enterprise Agent Platform, and Gemini Spark inside the consumer Gemini app for AI Pro and Ultra subscribers across more than 160 countries. That is a notably broad simultaneous rollout for a mid-cycle model refresh, and it reinforces the coding-and-agents positioning: developer surfaces (API, Antigravity, Android Studio) got it on day one alongside the consumer product.

Coding and agentic workflows

This is where Google concentrated the release. FrontierCode 1.1 Main - a production code-quality benchmark spanning around 100 programming tasks across multiple languages, scored on bug testing and style-guide compliance - improved from 34.4% to 43.6%, the single largest relative jump on Google's table and enough to lead every model listed, including GPT-5.6 Terra. Code Arena, a web-development-focused Elo ranking, moved from 1538 to 1588, again the top score on the table.

Enterprise document and knowledge work

On GDP.pdf, a benchmark testing expert-level comprehension of business PDF documents, Gemini 3.7 Flash scored 34.0% - roughly 6 points ahead of Claude Sonnet 5 and 9 points ahead of GPT-5.6 Terra on the same test. Harvey LAB-AA, which evaluates complex legal-workflow reasoning, put it at 90.7%, again the best result on the table and just ahead of Claude Sonnet 5's 90.1%. Long-context recall (GDM-MRCR v2, averaged at a 128K-token window) hit 97.0%, the strongest score of the five models compared.

Where it doesn't lead

Agentic computer-use (OSWorld-2.0) came in at 47.9%, behind GPT-5.6 Terra's 50.2%. On Agent's Last Exam - a pass-rate benchmark for multimodal desktop and OS agent tasks - Gemini 3.7 Flash scored 26.3%, behind both Claude Sonnet 5 (33.3%) and GPT-5.6 Terra (28.0%). And on GDPVal-AA v2, an Elo-style knowledge-work ranking, it placed behind Claude Sonnet 5, GPT-5.6 Terra and Muse Spark 1.2 alike, with Muse Spark 1.2 topping that particular table at 1628.

Rather than cherry-pick, here is the full table Google itself chose to publish, reproduced as prose. Bold in the source image marks the best score per row:

• Artificial Analysis Intelligence Index: 56 - behind GPT-5.6 Terra and Muse Spark 1.2 (both 57), ahead of Gemini 3.6 Flash (52) and just above Claude Sonnet 5 (55).
• FrontierCode 1.1 Main: 43.6% - best of the five, ahead of Claude Sonnet 5 (42.7%) and GPT-5.6 Terra (41.3%).
• DeepSWE v1.1: 65.3% - second-best; GPT-5.6 Terra leads at 69.6%.
• Code Arena Elo: 1588 - best of the five.
• Terminal-Bench 2.1: 85.8% and Terminal-Bench 3.0: 14.9% - both second-best behind GPT-5.6 Terra (87.4% and 20.8% respectively).
• AutomationBench (private set): 30.4% - best of the five, ahead of GPT-5.6 Terra (23.6%).
• GDPVal-AA v2 Elo: 1525 - lowest of the five; Muse Spark 1.2 leads at 1628.
• Harvey LAB-AA: 90.7% - best of the five, narrowly ahead of Claude Sonnet 5 (90.1%).
• GDP.pdf: 34.0% - best of the five.
• CharXiv Reasoning (no tools): 84.5% and (with tools): 88.7% - both trail GPT-5.6 Terra (85.9%, no-tools) and Gemini 3.6 Flash (89.4%, with-tools) respectively.
• LVBench (long video): 85.4% - best of the five.
• GDM-MRCR v2, 128K average: 97.0% - best of the five.
• OSWorld-2.0: 47.9% - behind GPT-5.6 Terra (50.2%); not tested for Claude Sonnet 5 or Muse Spark 1.2.
• Agent's Last Exam: 26.3% - lowest of the three models tested; Claude Sonnet 5 leads at 33.3%.
• HLE-Verified: 53.6% - best of the three models tested.
• BioMysteryBench (human-solvable): 87.1% - behind Claude Sonnet 5 (87.5%); (human-difficult): 43.5% - behind GPT-5.6 Terra (49.4%).
• LABBench2: 82.1% - best of the four models tested.

Read as a whole: Gemini 3.7 Flash is genuinely strong on production coding quality, web-dev Elo, long-context recall, and document/legal comprehension. GPT-5.6 Terra is genuinely stronger on the composite intelligence index, deep software-engineering (DeepSWE), terminal-agent tasks, and agentic computer use. Neither Google's launch framing nor a competitor's counter-framing captures that split accurately on its own - the table above is the more useful artefact than either headline.

Google DeepMind evaluated Gemini 3.7 Flash against its Frontier Safety Framework (April 2026 version) ahead of launch. Per the published model card, the model did not reach a tracked or critical capability level (CCL) in any assessed domain:

• CBRN (chemical, biological, radiological, nuclear): DeepMind's own assessment states the model "lacks the nuanced expert knowledge and actionable depth necessary to complete priority harm journeys" - the CCL threshold was not reached.
• Cybersecurity: the model reached DeepMind's internal alert threshold - a lower bar than the critical capability threshold - which triggered updated cyber-offence mitigations being deployed at launch as a precaution, rather than because a hard capability limit was actually crossed.
• Harmful manipulation: performance fell beneath the CCL alert threshold.
• ML R&D / misalignment: DeepMind reports the model cannot bypass its own testing restrictions and lacks the independence needed for end-to-end autonomous research workflows.

Specialist red-teaming found safety performance similar to or improved versus Gemini 3.6 Flash, and child-safety evaluations met Google's launch thresholds. The alert-threshold cyber result is the one figure in this section worth flagging clearly: it is not evidence of a dangerous capability that shipped anyway, but it is the closest this model comes to any tracked threshold, and it is why updated cyber mitigations specifically - rather than a blanket capability restriction - accompanied the release.

Unlike some competing labs, Google's Gemini 3.7 Flash materials do not include a dedicated hallucination-rate or calibration metric comparable to, say, Artificial Analysis's AA-Omniscience non-hallucination score used to evaluate other 2026 frontier models. The model card acknowledges, in general terms, that the model "exhibits hallucinations" and remains susceptible to jailbreak attempts, and notes occasional slowness or timeout issues in production use - but does not attach a quantified rate to any of that.

That is a real gap rather than a minor omission: HLE-Verified (53.6%) and Agent's Last Exam (26.3%) measure whether the model gets the right answer, not whether it knows when it doesn't know the answer. Buyers who need a specific honesty or calibration figure for risk-assessment purposes will not find one in Google's own launch materials for this release, and should treat that as an open question rather than assume parity with competitors that do publish the figure.

Launch-day coverage was broadly positive but split on emphasis. SiliconANGLE's coverage led with the coding and agent-ensemble angle and reproduced Google's claim that the model "outperformed comparable models from Anthropic and OpenAI across nine benchmarks" largely at face value. 9to5Google's reporting focused on the unusually fast three-week release cadence and Google's own explanation - developer feedback and algorithmic innovation - for why a full model generation wasn't needed to justify the update. Coverage consistently noted that Gemini 3.7 Flash is rolling out to Spark inside the Gemini app ahead of any confirmed Gemini 3.5 Pro release, an unusual sequencing worth watching as the wider Gemini 3 generation continues to fill out.

The nine-benchmarks framing holds up under scrutiny for what it claims - Gemini 3.7 Flash does lead on nine of the eighteen listed metrics against at least one of GPT-5.6 Terra or Claude Sonnet 5 - but neither outlet's coverage foregrounded that GPT-5.6 Terra leads the composite Intelligence Index and several individual coding and agentic benchmarks on the very same table Google published. Readers relying solely on the announcement's headline claim would reasonably conclude Gemini 3.7 Flash is the strongest model in this comparison set outright, which the full table does not support.

The detail worth catching: Gemini 3.7 Flash's introductory pricing is identical to Gemini 3.6 Flash's launch price, not a discount on it, despite several outlets characterising the release as coming at "half price." The 50% figure quoted in coverage refers to the gap between today's introductory rate and the standard rate that takes effect on 1 January 2027 - it is not a price cut relative to the previous model. Against Claude Sonnet 5 and GPT-5.6 Terra, Gemini 3.7 Flash undercuts both by a wide margin on a per-token basis at introductory pricing, which is a genuine part of its value case regardless of the benchmark nuance above.

• Not a clean sweep: GPT-5.6 Terra leads the Artificial Analysis Intelligence Index, DeepSWE v1.1, both Terminal-Bench versions, and OSWorld-2.0 on Google's own comparison table.
• No published hallucination or calibration metric for this release, unlike some competing labs' disclosures - the model card only acknowledges hallucinations exist in general terms.
• Introductory pricing is not a price cut: it matches Gemini 3.6 Flash's launch price rather than undercutting it, and both step up to the same standard rate on 1 January 2027.
• Cyber-safety alert threshold reached: not a critical capability breach, but the closest this model comes to any Frontier Safety Framework threshold, prompting updated mitigations at launch.
• Thin architectural disclosure: the model card defers most technical detail to the Gemini 3.6 Flash documentation rather than describing 3.7 Flash's changes directly.
• Ships ahead of its own generation's flagship: Gemini 3.5 Pro and Gemini 4 remain unreleased, so this is not yet a full picture of what the Gemini 3.x/4 generation can do at the top end.

Against GPT-5.6 Terra, Gemini 3.7 Flash trails on the composite Intelligence Index and on deep software-engineering and terminal-agent benchmarks, but leads on production code quality (FrontierCode), web-dev Elo, document comprehension and long-context recall - and costs meaningfully less at introductory pricing ($0.75/$3.75 vs $2/$12 per million tokens). Against Claude Sonnet 5, Gemini 3.7 Flash leads on most of the coding and document benchmarks but loses on Agent's Last Exam and GDPVal-AA v2, and again undercuts it substantially on price. Against its own predecessor, Gemini 3.6 Flash, the improvement is unambiguous and broad-based across nearly every listed benchmark, at unchanged introductory pricing.

The more useful framing than "which model wins" is task-fit: Gemini 3.7 Flash's specific strengths - production-quality code generation, long-document and legal-document comprehension, very long-context recall - line up well with coding-agent and enterprise-document workloads specifically, which is exactly how Google positioned it. For deep, multi-step software-engineering agentic work or terminal-heavy automation, GPT-5.6 Terra's lead on DeepSWE and Terminal-Bench is real and worth weighing against the price difference.

Choose Gemini 3.7 Flash if you're building coding agents, web-development tooling, long-document or legal-document processing pipelines, or any workload that benefits from a 1-million-token context window at a low per-token price - particularly if you're already inside Google's ecosystem via Antigravity, AI Studio or the Gemini Enterprise Agent Platform.

Look elsewhere if your workload is deep, multi-step software-engineering or terminal-agent work where GPT-5.6 Terra's benchmark lead is directly relevant, you need a disclosed hallucination or calibration figure for risk-assessment purposes, or you specifically need the strongest available composite reasoning score rather than task-specific strength.

Gemini 3.7 Flash is a genuinely capable, fast-turnaround update that improves on Gemini 3.6 Flash almost everywhere while holding its introductory price flat - a straightforwardly good deal against its own predecessor. Measured against outside competitors, it's a mixed but favourable picture: real, sourced leads on production code quality, document comprehension and long-context recall, and real, sourced losses on the composite intelligence index and deep agentic-coding benchmarks against GPT-5.6 Terra.

The Frontier Safety Framework results are reassuring rather than remarkable - no tracked capability level reached, with the cyber-alert threshold as the one figure worth watching going forward. For coding agents and document-heavy enterprise workloads specifically, Gemini 3.7 Flash is a strong, cheap default. For workloads that live and die on deep agentic software-engineering performance, GPT-5.6 Terra's benchmark lead on that specific axis is worth the extra cost.