Claude Sonnet 5vsGemini 3.5 Flash-Lite
Claude Sonnet 5 | Gemini 3.5 Flash-Lite | |
|---|---|---|
| API pricingUSD per 1M tokens · lower wins | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | $2.00 | $0.30 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | $10.00 | $2.50 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | $0.20 | $0.03 |
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $2.00Amazon Bedrock | — |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | $10.00Amazon Bedrock | — |
| Benchmarks | ||
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 80% | 65% |
SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better. | 63.2% | 54.2% |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 80.4% | 54% |
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better. | 81.2% | 74% |
| BenchmarksPublished by one model only | ||
Gray Swan IPI · k = 1Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. Gray Swan's indirect prompt injection benchmark measures how often such an attack succeeds when the attacker gets a single try. Lower is better. | 0.6% | — |
Gray Swan IPI · k = 10Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 10 tries and counts an attack as successful if any of them works. Lower is better. | 4.7% | — |
Gray Swan IPI · k = 15Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 15 tries and counts an attack as successful if any of them works. Lower is better. | 5.9% | — |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 85.2% | — |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | 81% | — |
Supabase Evals · with skillsSupabase coding — Supabase's own open benchmark: a coding agent is dropped into a real Supabase project and asked to do real work — set up a schema, fix a broken security policy, debug an Edge Function — and every run is checked against a live Supabase stack. This is the headline number, where the agent has Supabase's own skills loaded, as most people building on Supabase would. The score is the share of scenarios it got right. Higher is better. | 88.4% | — |
Supabase Evals · no skillsSupabase coding — The same Supabase scenarios, but with none of Supabase's skills loaded — so it measures what the model already knows about building on Supabase, rather than how well it follows Supabase's supplied instructions. Higher is better. | 81.2% | — |
Frontier-Bench v0.1Agentic computer work — A hard, ever-evolving set of real computer tasks — coding, system administration, data work, and more — that an AI agent has to complete on its own. Run by the Harbor / Laude Institute team as the successor to Terminal-Bench (v0.1 is the first release of the task set). The score is the share of tasks solved. Higher is better. | 14.6% | — |
Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. | 12.42% | — |
BU BenchBrowser agent — Can the AI drive a real web browser to finish tasks — clicking, filling forms, and navigating sites the way a person would? Run by Browser Use on their BU Bench task set. Higher is better. | — | 49% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 84.7% | — |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | 43.2% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 57.4% | — |
GDPval-AAKnowledge work — Measures how well the AI does economically valuable knowledge work, judged against human experts. Shown as a rating (like a chess Elo) — higher is better. | 1618 | — |
GDPval-AA v2Knowledge work — economically valuable knowledge work (v2, re-based Elo) | — | 1140 |
| Overview | ||
| Company | Anthropic | |
| Release date | Jun 30 2026 | Jul 21 2026 |
| Access | Proprietary | Proprietary |
Other comparisons
Claude Sonnet 5vsGPT-6 AstraGemini 3.5 Flash-LitevsGPT-6 AstraClaude Sonnet 5vsMuse Spark 1.3Gemini 3.5 Flash-LitevsMuse Spark 1.3Claude Sonnet 5vsGrok 4.6Gemini 3.5 Flash-LitevsGrok 4.6Claude Sonnet 5vsDeepSeek-V4-Pro-0813Gemini 3.5 Flash-LitevsDeepSeek-V4-Pro-0813Claude Sonnet 5vsMistral Medium 3.5Gemini 3.5 Flash-LitevsMistral Medium 3.5Claude Sonnet 5vsKimi K3Gemini 3.5 Flash-LitevsKimi K3
Frequently asked questions
Claude Sonnet 5 leads Gemini 3.5 Flash-Lite on 4 of the 4 benchmarks they both report (BullshitBench v2, SWE-Bench Pro, Terminal-Bench 2.1, OSWorld-Verified). Gemini 3.5 Flash-Lite is cheaper on both input and output: $0.30 vs $2.00 per million input tokens, and $2.50 vs $10.00 per million output tokens. Claude Sonnet 5 shipped 21 days before Gemini 3.5 Flash-Lite, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On BullshitBench v2, Claude Sonnet 5 leads at 80% vs Gemini 3.5 Flash-Lite at 65%. On SWE-Bench Pro, Claude Sonnet 5 leads at 63.2% vs Gemini 3.5 Flash-Lite at 54.2%. On Terminal-Bench 2.1, Claude Sonnet 5 leads at 80.4% vs Gemini 3.5 Flash-Lite at 54%. On OSWorld-Verified, Claude Sonnet 5 leads at 81.2% vs Gemini 3.5 Flash-Lite at 74%.
Claude Sonnet 5 was released by Anthropic on Jun 30 2026.
Gemini 3.5 Flash-Lite was released by Google on Jul 21 2026.
Claude Sonnet 5 leads on SWE-Bench Pro — Claude Sonnet 5 63.2% vs Gemini 3.5 Flash-Lite 54.2%.
Gemini 3.5 Flash-Lite is cheaper on both input and output: $0.30 vs $2.00 per million input tokens, and $2.50 vs $10.00 per million output tokens. Rates are pay-as-you-go API prices verified on August 18, 2026.