RankingGemini 4 Argon
Gemini 4 Argon
Gemini 4 Argon by Google is a Gemini 4 model, released 2026-09-30. This profile collects the published benchmark results and reported effort settings for this model. Configurations are compared on the effort page. Ranks are on the leaderboard.
30 published benchmark measures · 58 measured effort results · 3 benchmark families contribute across 2 task areas. 2 capability estimates available in the full profile. See all results ↓ · Compare published benchmarks →
Model evidence summary
This profile combines published settings. It is not a runnable configuration or a leaderboard rank. Compare measured configurations →
Performance profile
Capabilities
Filled points are supported; hollow points are preliminary. Lines stop at unknown capabilities. Exact values and sources follow below.
Scores estimate outcomes against a shared reference panel; they are not accuracy percentages. Sparse or disconnected evidence cannot qualify an overall profile. Open a capability to inspect its evidence.
Reported effort · Not specified
Settings reported in this model's published benchmark results, including results outside the aggregate. Effort names are provider-specific. These are not API defaults or equal compute budgets.
- Not specified: 32 observations
Mixed settings means multiple settings occur in the evidence. Best across efforts means the source selected its best reported result across settings; it does not mean Max. Unspecified settings stay unknown. This model-summary chart combines reported settings. The leaderboard keeps identified configurations separate and excludes unknown effort.
Inspect each result and its source ↓ · Download effort evidenceScore contributions and missing evidence
3 contributing families across 2 capabilities. Fixed reference panels do not change when the catalog expands.
Capability is fitted jointly across families. Capability estimates below describe different task areas; their weighted sum is not the Capability score.
Results without reviewed compatibility or a reference match remain in the raw evidence below. Coverage counts only contributing results.
Model information & shareable badge
- Lab
- Catalog status
- active
- Availability
- Globally available in the paid Gemini API in AI Studio and the Gemini platform, for Google AI Ultra subscribers in Google Antigravity, and for Google AI Ultra and Pro subscribers in the Gemini app and AI Mode.
- Family
- Gemini 4
- Released
- 2026-09-30
- Context
- —
- API list price
- $2 input / $10 output per million tokens
- License
- proprietary
- Model card
- https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/
- Default Capability family coverage
/badge/gemini-4-argon.svg
Gemini 4 Argon measured results by effort setting
Each row is one published result for one effort setting. Evidence collected 2026-11-22. Compare the effort settings →
| Benchmark | Effort | Score | Source |
|---|---|---|---|
| AA-AnalystAgent | High | 48.8% | Artificial Analysis |
| AA-Briefcase Elo | High | 1485 | Artificial Analysis |
| AA-Briefcase Elo | Medium | 1467 | Artificial Analysis |
| AA-LCR v1.1 | High | 79.7% | Artificial Analysis |
| AA-LCR v1.1 | Medium | 78.3% | Artificial Analysis |
| AA-Omniscience Accuracy | High | 48.3% | Artificial Analysis |
| AA-Omniscience Accuracy | Medium | 49.0% | Artificial Analysis |
| AA-Omniscience Index | High | 46.1 | Artificial Analysis |
| AA-Omniscience Index | Medium | 41.1 | Artificial Analysis |
| ARC-AGI-2 | High | 91.3% | ARC Prize |
| ARC-AGI-2 | Low | 42.1% | ARC Prize |
| ARC-AGI-2 | Medium | 87.1% | ARC Prize |
| ARC-AGI-3 · provider-adapter | High | 97.3% | ARC Prize |
| ARC-AGI-3 · provider-adapter | Low | 97.5% | ARC Prize |
| ARC-AGI-3 · provider-adapter | Medium | 97.3% | ARC Prize |
| ARC-AGI-3 · standard | Low | 49.8% | ARC Prize |
| ARC-AGI-3 · standard | Medium | 46.3% | ARC Prize |
| AutomationBench-AA | High | 77.5% | Artificial Analysis |
| AutomationBench-AA | Medium | 75.8% | Artificial Analysis |
| CritPt | High | 27.1% | Artificial Analysis |
| CritPt | Medium | 23.1% | Artificial Analysis |
| DAYJOB Finance | High | 20.3% | Surge AI |
| DAYJOB Healthcare | High | 9.6% | Surge AI |
| FrontierCode 1.1 · extended-devin-cli | High | 59.4% | Cognition |
| FrontierCode 1.1 · extended-devin-cli | Low | 60.6% | Cognition |
| FrontierCode 1.1 · extended-devin-cli | Medium | 60.5% | Cognition |
| FrontierCode 1.1 · main-devin-cli | High | 45.9% | Cognition |
| FrontierCode 1.1 · main-devin-cli | Low | 47.2% | Cognition |
| FrontierCode 1.1 · main-devin-cli | Medium | 47.5% | Cognition |
| FrontierCS 2 | Unspecified | 54.2% | FrontierCS |
| FrontierCS 2 | Unspecified | 39.6% | FrontierCS |
| FrontierSWE v2 | Unspecified | 55.0% | Proximal |
| GDP.pdf: All-pass Rate | High | 21.8% | Artificial Analysis |
| GDP.pdf: All-pass Rate | Medium | 25.6% | Artificial Analysis |
| GDPval-AA v2.1 Leaderboard | High | 1634 | Artificial Analysis |
| GDPval-AA v2.1 Leaderboard | Medium | 1578 | Artificial Analysis |
| Harvey LAB-AA v1.1: Hallucination-Gated All-Pass Rate | High | 3.9% | Artificial Analysis |
| Humanity's Last Exam | High | 57.1% | Artificial Analysis |
| Humanity's Last Exam | Medium | 49.9% | Artificial Analysis |
| LMArena Agent Arena | High | 7.6 | LMArena |
| MLCR-AA | High | 13.9% | Artificial Analysis |
| MMMU-Pro | Medium | 88.3% | Artificial Analysis |
| Multi-step SWE | High | 10.1% | Klavis AI |
| SciCode | High | 61.8% | Artificial Analysis |
| SciCode | Medium | 58.9% | Artificial Analysis |
| sudo L7 | High | 18.9% | Surge AI |
| SWE-Bench Pro V2 Hard · mini-swe-agent | High | 92.2% | Scale |
| Terminal-Bench v4.0 | High | 57.1% | Artificial Analysis |
| Terminal-Bench v4.0 | Medium | 55.1% | Artificial Analysis |
| Terminal-Bench-Science 0.1 | High | 52.9% | Artificial Analysis |
| Terminal-Bench-Science 0.1 | Medium | 48.6% | Artificial Analysis |
| Vals AI · Code Migration v1 | High | 68.2% | Vals AI |
| Vals AI · Excel Modeling Benchmark v1 | High | 75.2% | Vals AI |
| Vals AI · Finance Agent v2 | High | 65.4% | Vals AI |
| Vals AI · Legal Research Bench v1 | High | 54.8% | Vals AI |
| Vals AI · Tax Agent Bench v1: Weighted Pass | High | 79.0% | Vals AI |
| Vals AI · Terminal-Bench 2.1 | High | 57.7% | Vals AI |
| Vals AI · Vibe Code Bench v1.1 | High | 91.9% | Vals AI |
Benchmark scores & sources
Original results, evaluation harnesses, and evidence behind this model.
| Benchmark | Bucket | Score | Harness | Evidence | Source-recorded date |
|---|---|---|---|---|---|
| Agent's Last Exam | Supporting evidence | 39.5%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent. Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 15: Agent's Last Exam / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| ARC-AGI-2 | Hard reasoning | 91.3% | ARC Prize verifiedcontributes to capability | official board | 2026-10-13 |
| AutomationBench | Agentic | 51.3%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Private test set; results sourced from the official Zapier public leaderboard. Provider-published launch claim, not independent measurement by Google for every row. Private test set; results sourced from the official Zapier public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 2: AutomationBench / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| AutomationBench-AA | Supporting evidence | 77.51346945899195% | Artificial Analysis AutomationBench-AA | official board | 2026-11-21 |
| Bug Hunt Bench | Supporting evidence | 38.76% | Bug Hunt Bench · high | official board | 2026-10-14 |
| Chartography | Multimodal | 71.6%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Without tools; results from the official Surge public leaderboard. Provider-published launch claim, not independent measurement by Google for every row. Without tools; results from the official Surge public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 17: Chartography / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| CWE-bench · 1 | Supporting evidence | 68%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Official public leaderboard; ranking uses pass@1 with pass@4 tiebreaks. Grid records pass@1 only. Provider-published launch claim, not independent measurement by Google for every row. Official public leaderboard; ranking uses pass@1 with pass@4 tiebreaks. Grid records pass@1 only. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 19: CWE-bench 1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| DAYJOB Finance | Supporting evidence | 30.5% | DAYJOB Finance · high | official board | 2026-11-20 |
| DAYJOB Healthcare | Supporting evidence | 23.6% | DAYJOB Healthcare · high | official board | 2026-11-20 |
| DeepSWE · 1.1 | Coding | 77.9%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed with mini-swe agent. Astra from the official public leaderboard; Fable and Opus from their system cards. Highest scoring thinking level per Datacurve. Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed with mini-swe agent. Astra from the official public leaderboard; Fable and Opus from their system cards. Highest scoring thinking level per Datacurve. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 5: DeepSWE 1.1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| FrontierCS 2 | Supporting evidence | 54.2% | Antigravity | official board | 2026-11-20 |
| FrontierSWE · 2 | Supporting evidence | 55%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Results sourced from Proximal's official public leaderboard. Provider-published launch claim, not independent measurement by Google for every row. Results sourced from Proximal's official public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Historical provider claim; secondary and superseded for current FrontierSWE comparisons by the independent Proximal FrontierSWE v2 mean@5 board at https://www.frontierswe.com/. Version-unspecified claims are not v2 results. Excluded from weighted Overall. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 6: FrontierSWE 2 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| FrontierSWE v2 | Supporting evidence | 55% | Proximal FrontierSWE v2 · Proximus · unspecified | official board | 2026-11-22 |
| GraphWalks · 256K to 1M BFS F1 | Supporting evidence | 84.2%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. All models self-computed. 200 problems with context length between 256K and 1M tokens. BFS F1, not pass rate. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed. 200 problems with context length between 256K and 1M tokens. BFS F1, not pass rate. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 14: GraphWalks 256K to 1M BFS F1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| GraphWalks · up to 128K BFS F1 | Supporting evidence | 99.7%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. All models self-computed. 650 items with context length up to 128K tokens. BFS F1, not pass rate. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed. 650 items with context length up to 128K tokens. BFS F1, not pass rate. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 13: GraphWalks up to 128K BFS F1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| Harvey's Legal Agent Benchmark · 1 | Supporting evidence | 19.6%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Results sourced from Vals AI. Provider-published launch claim, not independent measurement by Google for every row. Results sourced from Vals AI. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 4: Harvey's Legal Agent Benchmark 1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| Humanity's Sixth Sense | Supporting evidence | 55.4% | HSS · highReported settingsFull 522-task pass@1, task mean of three attempts; every rubric criterion must pass. 95% task-cluster bootstrap CI. Claude-Opus-5 judge. Scale AI + Elorian. Unversioned board; source hashes identify captures. Display only. Excluded from weighted Overall. Source-recorded date is the capture date, not a proven evaluation date. | official board | 2026-10-31 |
| LABBench · 2 | Supporting evidence | 88.8%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. All models self-computed with Linux terminal, pre-installed bioinformatics tools, Python, R and internet access. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed with Linux terminal, pre-installed bioinformatics tools, Python, R and internet access. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 11: LABBench 2 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| LMArena Agent Arena | Supporting evidence | 10.31 | LMArena Agent Arena · high | official board | 2026-11-18 |
| LMArena Text Arena | Human pref | 1525 | LMArena Text | official board | 2026-10-13 |
| LVBench | Multimodal | 91.7%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. All models self-computed without tools. Gemini 1 FPS; Astra 800 frames, Fable 300 frames and Opus 600 frames due to API limits. Unequal frame budgets are excluded from matched comparison. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed without tools. Gemini 1 FPS; Astra 800 frames, Fable 300 frames and Opus 600 frames due to API limits. Unequal frame budgets are excluded from matched comparison. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 18: LVBench / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| Multi-step SWE | Supporting evidence | 10.1% | mini-swe-agent | official board | 2026-11-06 |
| OSWorld · 2.0 offline partial score | Agentic | 69.2%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed, maximum of three single-attempt runs; offline subset partial score. Official Docker/evaluator, 1080p, 500 steps, Gemini CUA harness, parallel batch tools, compaction, pyautogui actuation, screenshot-only observations, UI-specific function declarations, safety filters, official 08.08 patch. Astra from its official blog. Anthropic combined online/offline results are excluded. Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed, maximum of three single-attempt runs; offline subset partial score. Official Docker/evaluator, 1080p, 500 steps, Gemini CUA harness, parallel batch tools, compaction, pyautogui actuation, screenshot-only observations, UI-specific function declarations, safety filters, official 08.08 patch. Astra from its official blog. Anthropic combined online/offline results are excluded. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 16: OSWorld 2.0 offline partial score / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| PostTrainBench · 1.1 | Supporting evidence | 45.3%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. All models self-computed; OpenCode harness; 10-hour budget on one NVIDIA H100 GPU. Weighted aggregate across four base models and seven benchmarks. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed; OpenCode harness; 10-hour budget on one NVIDIA H100 GPU. Weighted aggregate across four base models and seven benchmarks. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 9: PostTrainBench 1.1 / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |
| RiemannBench | Supporting evidence | 76%Reported settings & sourceGemini API highest thinking settings; exact named effort not specified by Google. Results sourced from the official Surge public leaderboard. Provider-published launch claim, not independent measurement by Google for every row. Results sourced from the official Surge public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 12: RiemannBench / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30 | Published configuration | lab self-report | Reviewed 2026-09-30 |