Gemini 3.5 Flash Lite vs Mercury 2
Updated October 21, 2026. 30 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Gemini 3.5 Flash Lite for coding: it scores 33.6 on the Vibe Coding Index against 0.0 for Mercury 2.
Pick Mercury 2 when output tokens cost 3.3x less, or when what matters is WeirdML or ALE-Bench, where it leads.
- Shared results
- 30
- Gemini 3.5 Flash Lite only
- 53
- Mercury 2 only
- 12
- Like-for-like categories
- 6 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
Gemini 3.5 Flash Lite
Gemini 3.5 Flash Lite scores 33.6 against 0.0.
One benchmark behind each score
Intelligence
AA Intelligence Index
Gemini 3.5 Flash Lite
Gemini 3.5 Flash Lite leads, 22.2 against 10.8.
Same board, both measured
Agentic work
AA Agentic Index
Gemini 3.5 Flash Lite
Gemini 3.5 Flash Lite leads, 14.3 against 1.6.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only Gemini 3.5 Flash Lite has a GPQA Diamond result, 83.8%.
Directional only
Knowledge
MMLU Pro
No call
Only Gemini 3.5 Flash Lite has a MMLU Pro result, 85.8%.
Directional only
Web design
Design Arena: websites
No call
Only Mercury 2 has a Design Arena: websites result, 1036.
Directional only
Long documents
Listed context window
Gemini 3.5 Flash Lite
Gemini 3.5 Flash Lite takes 1.05M tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Mercury 2
Gemini 3.5 Flash Lite $0.0015, Mercury 2 $0.0006 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Gemini 3.5 Flash Lite has the number, $0.41.
Directional only
At a glance
- Coding index
- Gemini 3.5 Flash Lite 33.6 from 1 benchmark, Mercury 2 0.0 from 1 benchmark.
- Widest category
- General, 17.7 points apart on average over 12 shared benchmarks.
- Widest single gap
- LMArena WebDev: React: Gemini 3.5 Flash Lite 1408 against Mercury 2 1160, 41.3 points apart on a 0 to 100 scale.
- Output price
- Gemini 3.5 Flash Lite $2.50 against Mercury 2 $0.75 per million output tokens, 3.3x apart.
- Context window
- Gemini 3.5 Flash Lite takes 1.05M tokens, Mercury 2 128K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Gemini 3.5 Flash Lite | Bars | Mercury 2 | Ties |
|---|---|---|---|---|
| Coding | 6 | 2 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 5 | 0 | ||
| Knowledge | 3 | 0 | ||
| General | 12 | 0 | ||
| All shared boards | 28 | 2 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index62
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| LMArena WebDev | 1435 | 1171 | Gemini 3.5 Flash Lite |
| AA Coding Index | 49.3 | 31.1 | Gemini 3.5 Flash Lite |
| LiveCodeBench (Vals AI) | 79.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| SWE-bench Verified (Vals AI) | 75.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Vibe Code Bench (Vals AI) | 37.2% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Code migration (Vals AI) | 6.2% | Not measured | Only Gemini 3.5 Flash Lite measured |
| ProgramBench (Vals AI) | 0.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Terminal-Bench 2.1 (Vals AI) | 50.2% | Not measured | Only Gemini 3.5 Flash Lite measured |
| CyberBench (Vals AI) | 59.1% | Not measured | Only Gemini 3.5 Flash Lite measured |
| SciCode | 41.3% | 37.7% | Gemini 3.5 Flash Lite |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 26.5% | Only Mercury 2 measured |
| WeirdML | 39.0% | 43.2% | Mercury 2 |
| ALE-Bench | 765.3 | 785.6 | Mercury 2 |
| LMArena Text: coding | 1453 | 1390 | Gemini 3.5 Flash Lite |
| LMArena Text: software and IT | 1451 | 1385 | Gemini 3.5 Flash Lite |
| LMArena WebDev: React | 1408 | 1160 | Gemini 3.5 Flash Lite |
| LMArena WebDev: HTML | Not measured | 1199 | Only Mercury 2 measured |
| LMArena WebDev: from an image | 1448 | Not measured | Only Gemini 3.5 Flash Lite measured |
| Neither model has a result on the other 20 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| AA Intelligence Index | 22.2 | 10.8 | Gemini 3.5 Flash Lite |
| Epoch Capabilities Index | 144.9 | Not measured | Only Gemini 3.5 Flash Lite measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| AA Agentic Index | 14.3 | 1.6 | Gemini 3.5 Flash Lite |
| Finance Agent (Vals AI) | 47.4% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Tax Agent Bench (Vals AI) | 7.2% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Legal Research Bench (Vals AI) | 13.9% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Excel Modeling Benchmark (Vals AI) | 43.2% | Not measured | Only Gemini 3.5 Flash Lite measured |
| τ²-Bench Telecom | Not measured | 70.8% | Only Mercury 2 measured |
| APEX-Agents | 29.3% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Neither model has a result on the other 32 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| GPQA Diamond | 83.8% | Not measured | Only Gemini 3.5 Flash Lite measured |
| MMMU Pro (Vals AI) | 83.6% | Not measured | Only Gemini 3.5 Flash Lite measured |
| GPQA Diamond (Artificial Analysis run) | 83.8% | 77.0% | Gemini 3.5 Flash Lite |
| Humanity's Last Exam (Artificial Analysis run) | 18.8% | 17.1% | Gemini 3.5 Flash Lite |
| GPQA Diamond (Epoch AI run) | 83.3% | Not measured | Only Gemini 3.5 Flash Lite measured |
| ARC-AGI | 53.5% | Not measured | Only Gemini 3.5 Flash Lite measured |
| ARC-AGI-2 | 10.3% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Chess puzzles | 21.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| Mystery game puzzles | 19.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: hard prompts | 1441 | 1362 | Gemini 3.5 Flash Lite |
| LMArena Text: expert | 1434 | 1358 | Gemini 3.5 Flash Lite |
| LMArena Text: hard prompts in English | 1440 | 1374 | Gemini 3.5 Flash Lite |
| LMArena Vision: diagrams | 1284 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Vision: homework | 1291 | Not measured | Only Gemini 3.5 Flash Lite measured |
| Neither model has a result on the other 7 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 71.1% | Not measured | Only Gemini 3.5 Flash Lite measured |
| FrontierMath, tiers 1 to 3 | 26.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| FrontierMath, tier 4 | 0.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| ProofBench | 13.0% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: math | 1434 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: mathematical work | 1433 | Not measured | Only Gemini 3.5 Flash Lite measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| MMLU Pro | 85.8% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LegalBench (Vals AI) | 84.1% | Not measured | Only Gemini 3.5 Flash Lite measured |
| CorpFin (Vals AI) | 60.7% | Not measured | Only Gemini 3.5 Flash Lite measured |
| TaxEval (Vals AI) | 72.6% | Not measured | Only Gemini 3.5 Flash Lite measured |
| MortgageTax (Vals AI) | 68.7% | Not measured | Only Gemini 3.5 Flash Lite measured |
| MedCode (Vals AI) | 43.5% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: science | 1451 | 1343 | Gemini 3.5 Flash Lite |
| LMArena Text: medicine and healthcare | 1438 | 1363 | Gemini 3.5 Flash Lite |
| LMArena Text: legal and government | 1431 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: business and finance | 1423 | 1349 | Gemini 3.5 Flash Lite |
| Neither model has a result on the other 7 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1036 | Only Mercury 2 measured |
| Design Arena: UI components | Not measured | 998 | Only Mercury 2 measured |
| Design Arena: game dev | Not measured | 985 | Only Mercury 2 measured |
| Design Arena: data viz | Not measured | 1009 | Only Mercury 2 measured |
| Design Arena: 3D | Not measured | 982 | Only Mercury 2 measured |
| Design Arena: all code categories | Not measured | 1022 | Only Mercury 2 measured |
| Design Arena: SVG | Not measured | 998 | Only Mercury 2 measured |
| Design Arena: ASCII art | Not measured | 990 | Only Mercury 2 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index120
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| Public Benefits Bench (Vals AI) | 51.8% | Not measured | Only Gemini 3.5 Flash Lite measured |
| SAGE (Vals AI) | 47.3% | Not measured | Only Gemini 3.5 Flash Lite measured |
| MedScribe (Vals AI) | 70.9% | Not measured | Only Gemini 3.5 Flash Lite measured |
| IFBench | Not measured | 69.8% | Only Mercury 2 measured |
| Long context reasoning (AA-LCR) | 76.0% | 43.7% | Gemini 3.5 Flash Lite |
| DTBench | 83.5% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMCA | 37.6% | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text | 1435 | 1354 | Gemini 3.5 Flash Lite |
| LMArena Text: instruction following | 1419 | 1328 | Gemini 3.5 Flash Lite |
| LMArena Text: creative writing | 1421 | 1286 | Gemini 3.5 Flash Lite |
| LMArena Vision | 1269 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: English | 1438 | 1378 | Gemini 3.5 Flash Lite |
| LMArena Text: non-English | 1427 | 1330 | Gemini 3.5 Flash Lite |
| LMArena Text: Chinese | 1468 | 1414 | Gemini 3.5 Flash Lite |
| LMArena Text: German | 1454 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: Japanese | 1426 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: Korean | 1407 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: French | 1451 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: Spanish | 1441 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: Russian | 1443 | 1304 | Gemini 3.5 Flash Lite |
| LMArena Text: Polish | 1437 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Text: multi-turn | 1445 | 1357 | Gemini 3.5 Flash Lite |
| LMArena Text: longer queries | 1435 | 1329 | Gemini 3.5 Flash Lite |
| LMArena Text: writing and language | 1428 | 1300 | Gemini 3.5 Flash Lite |
| LMArena Text: entertainment and media | 1414 | 1314 | Gemini 3.5 Flash Lite |
| LMArena Vision: English | 1269 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Vision: Chinese | 1311 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Vision: reading text | 1280 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Vision: humor | 1287 | Not measured | Only Gemini 3.5 Flash Lite measured |
| LMArena Vision: creative writing | 1260 | Not measured | Only Gemini 3.5 Flash Lite measured |
| Neither model has a result on the other 16 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemini 3.5 Flash Lite | Mercury 2 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Mercury 2 costs less
Repository review
50K input + 3K output tokens
Mercury 2 costs less
Agent loop
200K input + 10K output tokens
Mercury 2 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Google Gemini 3.5 Flash Lite | Inception Mercury 2 |
|---|---|---|
| Maker | Inception | |
| Weights | No public weights listed | No public weights listed |
| Released | July 21, 2026 | March 4, 2026 |
| Context window | 1.05M | 128K |
| Input $/Mtok | $0.30 | $0.25 |
| Output $/Mtok | $2.50 | $0.75 |
| Coding index | 33.6 | 0.0 |
| Independent measurements behind it | 1 | 1 |
| Cost per solved task | $0.41 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemini 3.5 Flash Lite better than Mercury 2 for coding?
Pick Gemini 3.5 Flash Lite for coding: it scores 33.6 on the Vibe Coding Index against 0.0 for Mercury 2. Pick Mercury 2 when output tokens cost 3.3x less, or when what matters is WeirdML or ALE-Bench, where it leads.
Which is cheaper, Gemini 3.5 Flash Lite or Mercury 2?
Gemini 3.5 Flash Lite costs $0.30 per million input tokens and $2.50 per million output tokens; Mercury 2 costs $0.25 and $0.75. Cost per solved task is known for Gemini 3.5 Flash Lite only.
Which has the bigger context window, Gemini 3.5 Flash Lite or Mercury 2?
Gemini 3.5 Flash Lite, with 1.05M tokens against 128K.
How many benchmarks is the Gemini 3.5 Flash Lite vs Mercury 2 comparison based on?
30 benchmarks have a result for both models as of October 21, 2026. Gemini 3.5 Flash Lite's coding index rests on 1 independent benchmark and Mercury 2's on 1. Where only one model has a result, the page says so instead of naming a winner.