Gemini 4 Neon Launches, Topping Coding Agent Index and Reshaping the Cost-Performance Frontier
Google released Gemini 4 Neon, its fourth Gemini 4 model in seven weeks and only 6 days after Hydrogen. The model immediately posted top scores across multiple benchmarks and reset the intelligence-cost frontier at a markedly lower price, with many calling it one of the strongest models of 2026.
Confirmed
- Leaderboard results: On the Artificial Analysis Coding Agent Index v2.1, Neon (Antigravity CLI, xhigh tier) took first place with 59 points — 3 points above Carbon (high, 55), tied with Claude Opus 5.5, roughly 5-6 points ahead of GPT-6.2 Astra, and ahead of Claude Fable 5.2 (confirmed across multiple posts by kimmonismus and Artificial Analysis)
- Bug Hunt Bench: Pawel Huryn tested it on 2 real repositories with 105 seeded bugs; Neon fixed 63 to top the board, narrowly beating GPT-6.2 Astra
- Agent Arena: Official data shows Neon (High) ranked #2 with 1336 points, behind only Claude Fable 5.2 (Max) at 1453; across 4.2k real agent sessions, median cost per task was $2.82, about 53% cheaper than Opus 5.5, resetting the Pareto frontier
- Intelligence Index: Ranked #3 on the Artificial Analysis Intelligence Index (WorldofAI's hands-on testing says it competes with Opus 5.5 and GPT-6.2); record Quality Elo of 1771 and AA-Briefcase total of 1700, up 89 points over Carbon but still behind Claude Fable
- Cost data: Priced at $3/$15, roughly half to 1/2.5 of Claude Fable ($8/$40); $6.44 per task (xhigh), about 17% below Carbon (high, $7.76), but about 1.6x GPT-6.2 Astra (max); Hydrogen is lowest at $0.29/task, with Neon and Hydrogen sweeping every Pareto-optimal point in the Gemini 4 lineup
- Release cadence: Arena One data shows DeepMind's release gaps shrinking exponentially: Argon to Carbon 28 days, Carbon to Hydrogen 15 days, Hydrogen to Neon 6 days
Why it matters
- kimmonismus judged that Google pulled off what OpenAI has been trying to do; WorldofAI called Neon one of the strongest models of 2026 — faster and cheaper
- Google engineer Melvin Johnson (relayed by Lenny Rachitsky) said Neon narrowly beat Opus 5.5 Max on Agents Arena; dummerspast39 argued its low price makes it viable for fleets of thousands of agents
- Artificial Analysis benchmarks show divergence within the Gemini 4 family: Carbon dominates the graduate-level physics benchmark CritPt but trails Neon and Hydrogen on agentic research and knowledge work, indicating each variant has its own focus
2026-11-19 ~ 2026-11-20 · 28 related posts
- Episode 1: Unreleased Gemini 4 Neon Appears in Google Cloud Billing, Priced $3/$15(2026-11-18, 7 posts)
- Episode 2: Google Launches Gemini 4 Neon: Frontier Intelligence at Fast-Reasoning Speed(2026-11-18, 96 posts)
- Episode 3: Gemini 4 Neon rolls out quietly with three-tier model lineup(2026-11-19, 4 posts)
- Episode 4: Gemini 4 Neon Launches, Topping Coding Agent Index and Reshaping the Cost-Performance Frontier(2026-11-19, 28 posts)
- Episode 5: Gemini 4 Neon's SWE-sweep Score Challenged as Harness-Dependent(2026-11-19, 2 posts)
- Episode 6: Gemini 4 Neon Tops CursorBench Pareto Frontier(2026-11-19, 2 posts)
- Episode 7: Gemini 4 Neon Wins Praise From Developers, Nearly Beating Fable 5.2 Across the Board(2026-11-19, 23 posts)
- Episode 8: Google's Gemini 4 Neon Tops Vending-Bench 2 Agent Benchmark(2026-11-19, 2 posts)
- Episode 9: Leaked Gemini 4 Neon Benchmark Chart Shows Leads Over Opus 5.5 and GPT-6.2(2026-11-19, 3 posts)
Primary sources
- Gemini 4 Neon (High) hits Agent Arena Pareto frontier at $2.82 median cost per task — arena ·
- Gemini 4 Neon hits 59 on Coding Agent Index, matching Claude flagships at half the cost — ArtificialAnlys ·
- Gemini 4 Neon tops Bug Hunt bench, fixing 63/105 real repo bugs to edge out GPT-6.2 Astra — PawelHuryn ·
- Gemini 4 Neon takes #2 on LMArena agent index at a fraction of Claude's cost — lyraxana · 2026-11-19
- [source] Gemini 4 Neon (High) hits Agent Arena Pareto frontier at $2.82 median cost per task — arena · 2026-11-19
- Gemini 4 Neon hits #2 on Agent Arena Index at 53% lower cost than Claude Opus 5.5 — arena · 2026-11-19
- Gemini 4 Neon matches Claude Opus 5.5 on coding agents at roughly half the cost — ArtificialAnlys · 2026-11-19
- Gemini 4 Neon sets record 1771 Elo on Presentation, still trails Claude on Briefcase — ArtificialAnlys · 2026-11-19
- Gemini 4 Neon costs $6.44 per task, 17% below Carbon but 1.6x GPT-6.2 Astra — ArtificialAnlys · 2026-11-19
- Gemini 4 Neon scores 61 on AA Index, beats Carbon, burns ~1.5% of 5-hour limit per task — Immediate_Simple_217 · 2026-11-19
- Google engineer says Gemini Neon edges out Opus 5.5 Max on Agents Arena — lennysan · 2026-11-19
- Gemini Neon beats Gemini Carbon on Artificial Analysis despite being cheaper — reddit4jonas · 2026-11-19
- Gemini 4 Neon High matches xHigh scores at 42% lower cost — adonis_singh · 2026-11-19
- Gemini 4 Neon takes #1 on Coding Agent Index at $3/$15, ending Anthropic's top-3 sweep — olcan · 2026-11-19
- Gemini 4 Neon tested: ranks #3 on Intelligence Index, takes on Opus 5.5 and GPT-6.2 — WorldofAI · 2026-11-19
- Gemini 4: Neon (low) matches or beats Hydrogen (medium) at lower cost — fundal_alb · 2026-11-19
- Surge AI benchmarks Gemini 4 Neon: tops Tuesday@Work at 71.6, trails on SWE — echen · 2026-11-19
- Gemini 4 Neon tops Coding Agent Index at 59, 6 points ahead of GPT-6.2 Astra — kimmonismus · 2026-11-19
- Gemini 4 Neon Tops Coding Agent Index, 6 Points Ahead of GPT-6.2 Astra — kimmonismus · 2026-11-19
- Gemini 4 Neon takes #1 on Coding Agent Index at $3/$15, 60% cheaper than Fable — clmt · 2026-11-19
- Gemini 4 Neon hits #2 on Agent Arena as DeepMind's release cadence shrinks to 6 days — lmthang · 2026-11-19
- Google's Gemini 4 Neon ranks #3 on intelligence index at less than half Claude's cost — AI寒武纪 · 2026-11-19
- [source] Gemini 4 Neon tops Bug Hunt bench, fixing 63/105 real repo bugs to edge out GPT-6.2 Astra — PawelHuryn · 2026-11-19
- Google launches Gemini 4 Neon, debuts No. 1 on coding agent index — tallinzen · 2026-11-19
- Google ships Gemini 4 Neon in 7 weeks' 4th release, tops coding agent index at $3/$15 — TheAIGRID · 2026-11-20
- Gemini 4 Carbon tops graduate-level physics benchmark but loses on agentic tasks — ArtificialAnlys · 2026-11-20
- Neon and Hydrogen own every Pareto frontier in Gemini 4 lineup, from $0.29 to $6.44 per task — ArtificialAnlys · 2026-11-20
4 near-duplicate retellings: ArtificialAnlys · melvinjohnsonp · rseroter · ArtificialAnlys