Compare / Claude Haiku 5.5 vs Gemini 3.5 Flash-Lite
Claude Haiku 5.5 vs Gemini 3.5 Flash-Lite
Which is cheaper, which is more capable, and which does better at each kind of work, from 91 results on 6 sources that measured both models.
Results as of 8 October 2026 (catalogue release 2026-10-08-4199ce0620fe).
Claude Haiku 5.5 or Gemini 3.5 Flash-Lite? The short answer
Pick Claude Haiku 5.5: it scores higher (Spring Prompt overall 44 against 27) and costs less. Pick Gemini 3.5 Flash-Lite only if speed matters more: it writes faster.
- Spring Prompt overall: Claude Haiku 5.5 44, Gemini 3.5 Flash-Lite 27 (out of 100; how the score works).
- Claude Haiku 5.5 is cheaper to run: $0.001 against $0.002 per product listing on CatalogBench (41% less).
- Claude Haiku 5.5 scores higher on the Artificial Analysis Intelligence Index: 43.4 against 22.2.
- Gemini 3.5 Flash-Lite writes faster: 335 tokens a second against 235.
- Claude Haiku 5.5 is newer: released 7 Oct 2026, against 21 Jul 2026 for Gemini 3.5 Flash-Lite.
- On our own benchmarks, on the results both models have: Claude Haiku 5.5 does better for product listings, decks from an analysis and user surveys; Gemini 3.5 Flash-Lite for marketing planning and speed under pressure.
- On third-party benchmarks, on the results both models have: Claude Haiku 5.5 does better for agents and tool use and reasoning and knowledge.
At a glance
| Claude Haiku 5.5 | Gemini 3.5 Flash-Lite | |
|---|---|---|
| Price per million tokens | $0.10 in · $0.50 out | $0.30 in · $2.50 out |
| Blended (3 in : 1 out) | $0.20 | $0.85 |
| Cheapest host, blended | – | $0.85 (Google) |
| Cost per product listing (CatalogBench) | $0.001 | $0.002 |
| Cached input, per million | – | $0.030 cached input |
| Intelligence Index | 43.4 | 22.2 |
| Spring Prompt overall | 44 | 27 |
| Output speed, tokens a second | 235 | 335 |
| Context window | 1,000,000 tokens | 1,048,576 tokens |
| Released | 7 Oct 2026 (newer) | 21 Jul 2026 |
| Weights | Closed | Closed |
| Developer based in | the US | the US |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Claude Haiku 5.5 | Gemini 3.5 Flash-Lite |
|---|---|---|
| Product listingsTurning a sparse product feed and photos into listings that can go live | 11 of 20 | 16 of 20 |
| Decks from an analysisTurning a finished analysis into a deck you could present as it is | 10 of 19 | 17 of 19 |
| User surveysPlanning a user survey and reading its results without being misled | 3 of 8 | 6 of 8 |
| Marketing planningPlanning a year of ad spend without overspending | 17 of 19 | 15 of 19 |
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 20 of 164 | 73 of 164 |
| Professional workReal tasks from banking, consulting and law, business documents and freelance projects | not measured | 46 of 51 |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 19 of 198 | 91 of 198 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | not measured | 49 of 188 |
| Sticking to the factsSummarising without inventing things, and factual answers | not measured | 62 of 152 |
| Speed under pressureGood decisions against a real clock (fast chess) | 15 of 22 | 2 of 22 |
From our research
Every shared result
Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.
| Measure | Claude Haiku 5.5 | Gemini 3.5 Flash-Lite |
|---|---|---|
| BulletBench measured by Spring Prompt · 3 results | ||
| Cost per game · Blitz 3+2US dollars, lower is better | $0.0164 | $0.0087 |
| Cost per game · Bullet 60sUS dollars, lower is better | $0.0019no reasoning | $0.0058 |
| Ladder Elo · Bullet 60sladder Elo, higher is better | 310no reasoning | 780 |
| CatalogBench measured by Spring Prompt · 3 results | ||
| Cost per productUS dollars, lower is better | $0.0012 | $0.0020 |
| Reliably publish-ready% of products, higher is better | 26.8% | 17.9% |
| Reliably publish-ready · sales brief% of products, higher is better | 10.7% | 3.6% |
| DeckBench measured by Spring Prompt · 3 results | ||
| Cost per deckUS dollars, lower is better | $0.0091 | $0.0126 |
| Deck ratingrating, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. | 1,130 | 645 |
| Accurate decks% of tasks, higher is better | 33.3% | 0.0% |
| ROASBench measured by Spring Prompt · 3 results | ||
| Overall scorescore out of 100, higher is better | 24.1 | 25.4 |
| Audience scorescore out of 100, higher is better | 48.4 | 45.0 |
| Business scorescore out of 100, higher is better | 13.5 | 12.6 |
| SurveyBench measured by Spring Prompt · 3 results | ||
| Analysis ratingrating, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. | 1,412 | 662 |
| Cost per taskUS dollars, lower is better | $0.0185 | $0.0144 |
| Analysis scorepoints of 100, higher is better | 88.4 | 56.7 |
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better | 43.4max reasoning | 22.2 |
| Humanity's Last Exam% of questions, higher is better | 44.4%max reasoning | 18.8% |
| Terminal-Bench 4.0% of tasks, higher is better | 32.8%max reasoning | 1.0% |
Show all 91 results (73 more)
| Measure | Claude Haiku 5.5 | Gemini 3.5 Flash-Lite |
|---|---|---|
| BulletBench measured by Spring Prompt · 7 results | ||
| Games lost on time · Blitz 3+2% of games, lower is better | 37.5% | 0.0% |
| Games lost on time · Bullet 60s% of games, lower is better | 37.5%no reasoning | 0.0% |
| Invalid moves · Blitz 3+2% of moves, lower is better | 0.0% | 0.5% |
| Invalid moves · Bullet 60s% of moves, lower is better | 0.0%no reasoning | 0.0% |
| Ladder Elo · Blitz 3+2ladder Elo, higher is better | 557 | 707 |
| Median move time · Blitz 3+2milliseconds, lower is better | 6.2 s | 0.8 s |
| Median move time · Bullet 60smilliseconds, lower is better | 1.7 sno reasoning | 0.9 s |
| CatalogBench measured by Spring Prompt · 27 results | ||
| Channel compliance% of products, higher is better | 100.0% | 98.2% |
| Channel rules broken% of products, lower is better | 0.6% | 1.8% |
| Channel rules broken · sales brief% of products, lower is better | 0.0% | 1.2% |
| Claims to check% of products, lower is better | 0.6% | 1.8% |
| Claims to check · sales brief% of products, lower is better | 1.2% | 7.7% |
| Conflicts caught% of conflicts, higher is better | 100.0% | 84.9% |
| Content quality% of checks, higher is better | 95.6% | 95.5% |
| Content quality · sales brief% of checks, higher is better | 96.3% | 94.8% |
| Decision accuracy% of decisions, higher is better | 95.6% | 93.4% |
| Failed outputs% of products, lower is better | 0.0% | 0.0% |
| Failed outputs · sales brief% of products, lower is better | 0.0% | 0.0% |
| Field accuracy% of missing fields, higher is better | 91.0% | 92.5% |
| Invented values% of filled values, lower is better | 0.8% | 0.8% |
| Missing UK information% of products, lower is better | 0.0% | 0.0% |
| Missing UK information · sales brief% of products, lower is better | 0.0% | 0.0% |
| Not findable% of products, lower is better | 11.9% | 20.2% |
| Not findable · sales brief% of products, lower is better | 10.7% | 21.4% |
| Publish-ready listings% of products, higher is better | 47.0% | 35.1% |
| Publish-ready listings · sales brief% of products, higher is better | 20.2% | 10.7% |
| Unsupported claims% of products, lower is better | 32.1% | 33.3% |
| Unsupported claimsclaims per product, lower is better | 1.20 | 1.18 |
| Unsupported claims · sales brief% of products, lower is better | 70.2% | 78.6% |
| Unsupported claims · sales briefclaims per product, lower is better | 2.46 | 3.44 |
| Wrong attributes% of products, lower is better | 19.6% | 34.5% |
| Wrong attributes · sales brief% of products, lower is better | 18.5% | 35.1% |
| Wrong category or variant% of products, lower is better | 0.0% | 9.5% |
| Wrong category or variant · sales brief% of products, lower is better | 0.0% | 10.7% |
| DeckBench measured by Spring Prompt · 13 results | ||
| Caveat dropped% of tasks, lower is better | 16.7% | 16.7% |
| Clean layout% of tasks, higher is better | 66.7% | 33.3% |
| Design quality% of the maximum, higher is better | 61.1% | 60.2% |
| Draft figure quoted% of tasks, lower is better | 33.3% | 50.0% |
| Findings missing% of tasks, lower is better | 16.7% | 33.3% |
| Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. | 50.0% | 26.4% |
| Layout defects% of tasks, lower is better | 33.3% | 66.7% |
| Misleading metric used% of tasks, lower is better | 16.7% | 0.0% |
| Presentable decks% of tasks, higher is better | 33.3% | 0.0% |
| Recommendation late or wrong% of tasks, lower is better | 33.3% | 83.3% |
| Slides needing work% of tasks, lower is better | 16.7% | 16.7% |
| Unsupported claims% of tasks, lower is better | 33.3% | 16.7% |
| Unsupported numbers% of tasks, lower is better | 16.7% | 0.0% |
| ROASBench measured by Spring Prompt · 6 results | ||
| Consistency scorescore out of 100, higher is better | 10.6 | 21.8 |
| Contribution profitsimulated US dollars, higher is better | $32,146 | $11,834 |
| Cost of a runUS dollars, lower is better | $0.0313 | $0.0390 |
| Months over budgetmonths of 12, lower is better | 0 | 0 |
| Planning scorescore out of 100, higher is better | 54.8 | 54.9 |
| Return on ad spendprofit per $1 spent, higher is better | $0.03 | $0.01 |
| SurveyBench measured by Spring Prompt · 16 results | ||
| Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other analysis was; the win rate counts wins against whichever analyses a model happened to meet, so the two can disagree. | 72.5% | 32.6% |
| Leading or double-barrelled questions% of tasks, lower is better | 33.3% | 77.8% |
| Mistaken next steps% of tasks, lower is better | 22.2% | 44.4% |
| Numbers right% of answers, higher is better | 100.0% | 75.5% |
| Objectives unanswerable% of tasks, lower is better | 0.0% | 33.3% |
| Precision ignored% of tasks, lower is better | 0.0% | 100.0% |
| Research scorepoints of 100, higher is better | 90.2 | 55.6 |
| Sound research% of tasks, higher is better | 11.1% | 0.0% |
| Stakeholder's view presumed% of tasks, lower is better | 0.0% | 55.6% |
| Survey plan scorepoints of 100, higher is better | 94.5 | 53.1 |
| Survey structure% of tasks, lower is better | 22.2% | 55.6% |
| Traps handled% of traps, higher is better | 94.2% | 56.5% |
| Traps missed% of tasks, lower is better | 22.2% | 100.0% |
| Unsupported findings% of tasks, lower is better | 77.8% | 66.7% |
| Wrong numbers% of tasks, lower is better | 0.0% | 100.0% |
| Wrong recommendation% of tasks, lower is better | 11.1% | 22.2% |
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| Long-context reasoning (AA-LCR)% of questions, higher is better | 82.7%max reasoning | 76.0% |
| Output speedtokens per second, higher is better | 235max reasoning | 335 |
| SciCode% of problems, higher is better | 55.0%max reasoning | 41.3% |
| Time to first answer tokenseconds, lower is better | 5.08low reasoning | 7.29 |
Quick answers
Is Claude Haiku 5.5 better than Gemini 3.5 Flash-Lite?
Pick Claude Haiku 5.5: it scores higher (Spring Prompt overall 44 against 27) and costs less. Pick Gemini 3.5 Flash-Lite only if speed matters more: it writes faster.
Which is cheaper, Claude Haiku 5.5 or Gemini 3.5 Flash-Lite?
Claude Haiku 5.5 is cheaper to run: $0.001 against $0.002 per product listing on CatalogBench (41% less).
Which is faster, Claude Haiku 5.5 or Gemini 3.5 Flash-Lite?
Gemini 3.5 Flash-Lite writes about 335 tokens a second on its usual API, against 235 for Claude Haiku 5.5 (Artificial Analysis).
Run Claude Haiku 5.5 and Gemini 3.5 Flash-Lite on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.