Get started

Gemini 3.5 Flash (batch)

Sold by 6 sellers, timed at 137 tok/s, scored 70.1 on the Coding Index, on 25 Design Arena boards, and with results on 2 openly licensed boards.

retrieved sha256 · 0bb53a77713f

Models/Google/Gemini 3.5 Flash (batch)USD per 1M tokens · as OpenRouter routes it
google/gemini-3.5-flash:batchGoogle · batch variant · Gemini tokenizer
OpenRouter list price · the routed rate, as the catalog carried itretrieved 17 Nov 16:29 UTC from openrouter.ai ↗
Input$0.75per 1M tokens
Output$4.50per 1M tokens
Cached input$0.075per 1M tokens
Context1.05M1,048,576 tokens
The other meters · every further rate the list publishes for this id, in the unit it states3 meters
Reasoning tokens$4.50per 1M tokens
Audio input$1.50per 1M tokens
Web search$0.014per search
What the list says · the fields OpenRouter publishes for this id12 fields
AuthorGoogle
Listed
Canonical idgoogle/gemini-3.5-flash-20260519
Context1,048,576 tokens
Max output65,536 tokens
Modalitiestext, image, video, file, audio → text
TokenizerGemini
Reasoningalways on · effort high, medium, low, minimal, medium by default
Moderationnot moderated by OpenRouter
Variantbatch
Knowledge cutoff2025-01-01
Other listingbase listing

What sellers charge

6 sellers · 3.6x from cheapest to dearest

One tick per seller, on a log scale of input price per million from $0.75 at Google to $2.70 at Google AI Studio. The $0.75 listed is the cheapest on offer; 4 sellers are dearer. A faint tick declares no number format. Every seller is named on the endpoint ledger below.

Measured speed

15,096 requests · 7 endpoints · 30-minute window
Throughput137 tok/sMedian output rate, weighted across the 7 endpoints serving it by the requests each answered.the speed board
First token1.39sMedian wait before the first token, over the same window and weighted the same way.how Vitna reads a figure
Endpoint spread64 to 162The slowest and the fastest route serving this model, in tokens per second. Speed is a property of the route, not of the weights.the board, fastest first

OpenRouter's own timing of its routed traffic, weighted by requests, over one 30-minute window read . Every route is on the endpoint ledger below.

Every endpoint

6 priced · 7 timed · cheapest first
  1. Googlegoogle-vertex/global/flex · listed$0.75$4.50$0.075not declared66knot matchednonenonenot matched78.94%
  2. Google AI Studiogoogle-ai-studio/flex · listed$0.75$4.50$0.075not declared66knot matchednonenonenot matched100%
  3. Googlegoogle-vertex/global$1.50$9.00$0.15not declared66knot matchednonenonenot matched100%
  4. Google AI Studiogoogle-ai-studio$1.50$9.00$0.15not declared66knot matchednonenonenot matched100%
  5. Googlegoogle-vertex/global/priority$2.70$16.20$0.27not declared66knot matchednonenonenot matched100%
  6. Google AI Studiogoogle-ai-studio/priority$2.70$16.20$0.27not declared66knot matchednonenonenot matched100%
  7. Googlenot matchednot matchednot matchednot matchedno offer64 tok/s1613s16 reqno offer
  8. Googlenot matchednot matchednot matchednot matchedno offer93 tok/s43.71.76s6 reqno offer
  9. Googlenot matchednot matchednot matchednot matchedno offer115 tok/s50.81.67s7 reqno offer
  10. Googlenot matchednot matchednot matchednot matchedno offer135 tok/s451.36s13,887 reqno offer
  11. Google AI Studionot matchednot matchednot matchednot matchedno offer70 tok/s491.34s1 reqno offer
  12. Google AI Studionot matchednot matchednot matchednot matchedno offer162 tok/s551.68s1,097 reqno offer
  13. Google AI Studionot matchednot matchednot matchednot matchedno offer99.5 tok/s581.4s82 reqno offer

Each track runs from a route's p90, the rate nine in ten of its requests met or beat, up to its median. Weighted by requests, the figure above reports where the traffic went rather than what one route does: 6 of the 7 timed routes are slower than it, and the middle route runs at 99.5 tok/s. The widest median-to-p90 gap is Google's, 4x.

Google publishes 3 offers and had 4 routes timed; Google AI Studio publishes 3 offers and had 3 routes timed. The two documents share no route id, so those offers and routes are printed apart rather than paired.

Uptime is OpenRouter's own figure for each seller over the trailing day, read once a day, and the column prints the latest reading, taken . 4 of the 7 sellers with a reading fell under 99% on at least one of the 23 days held, 23 Oct 2026 to 21 Nov 2026: Google (google-vertex/global/flex) at 0.00% on 6 Nov 2026, Google AI Studio (google-ai-studio/priority) at 75.00% on 15 Nov 2026, Google (google-vertex/global/priority) at 95.00% on 23 Oct 2026, and Google AI Studio (google-ai-studio) at 98.29% on 19 Nov 2026. Source: OpenRouter's public endpoints API, one reading a day for each seller: 1,211 sellers carried one on the latest of the 23 days held.

Source: OpenRouter's public endpoints API. 1257 sellers across 359 models when the mirror ran, 186 of them served by more than one. List prices, not what anybody paid; a discount is the seller's own. Read . Speed is OpenRouter's own timing over one 30-minute window, read .

What each seller accepts

12 parameters · 6 sellers
  1. include_reasoningevery seller
  2. max_tokensevery seller
  3. reasoningevery seller
  4. reasoning_effortevery seller
  5. response_formatevery seller
  6. seedevery seller
  7. structured_outputsevery seller
  8. temperatureevery seller
  9. tool_choiceevery seller
  10. toolsevery seller
  11. top_pevery seller
  12. stop3 of 6not Google AI Studio (google-ai-studio), Google AI Studio (google-ai-studio/flex), and Google AI Studio (google-ai-studio/priority)

11 of these 12 are accepted by every one of the 6 sellers that publish a list; the other 1 are declared by some sellers and not others, so what a request sending one of them gets depends on which seller serves it. Each list is the seller's own declaration in OpenRouter's endpoints API, read , and nothing here tested one.

Price position

342 priced models

Blended $1.688 per million tokens, 3 in to 1 out: cheaper than 113 of the 342 priced models in the catalog. Each bar is one stretch of the catalog on a logarithmic price scale; the mark is this listing.

Measured quality

145 models with an index · 132 on a design arena board

Each mark is one board this model stands on, placed at the share of that board it scores above, and led to its number in the list under the rule. The columns and the ledger below carry the figures.

Intelligence Indexnot scoredArtificial Analysis has not scored this model on its Intelligence Index.the intelligence leaderboard
Coding Index70.1Position 25 of 118, above 79% of the models scored.the coding leaderboard
Agentic Index26.0Position 33 of 94, above 65% of the models scored.the agentic leaderboard
  1. Agentic game development · Agents1126 Elowin rate 55.3% · source rank 13the board, 25 ranked
  2. Agentic HTML slides · Agents1162 Elowin rate 45.8% · source rank 6the board, 9 ranked
  3. Agentic slides · Agents1244 Elowin rate 57.5% · source rank 3the board, 9 ranked
  4. Agentic slides, HTML · Agents1162 Elowin rate 45.7% · source rank 6the board, 9 ranked
  5. Agentic slides, python-pptx · Agents1242 Elowin rate 57.8% · source rank 2the board, 9 ranked
  6. Android native · Agents1174 Elowin rate 57.2% · source rank 20the board, 41 ranked
  7. blender3d · Agents904 Elowin rate 31.3% · source rank 24the board, 25 ranked
  8. codetovideo · Agents1168 Elowin rate 50.2% · source rank 16the board, 26 ranked
  9. Full stack · Agents1170 Elowin rate 55.8% · source rank 22the board, 44 ranked
  10. Godot game development · Agents1069 Elowin rate 43.0% · source rank 31the board, 34 ranked
  11. HTML slides · Agents1040 Elowin rate 46.6% · source rank 21the board, 25 ranked
  12. Mobile apps · Agents1169 Elowin rate 50.0% · source rank 29the board, 49 ranked
  13. pdf · Agents1148 Elowin rate 46.2% · source rank 28the board, 36 ranked
  14. PPTX slides · Agents1244 Elowin rate 57.7% · source rank 2the board, 8 ranked
  15. python-pptx slides · Agents1247 Elowin rate 57.4% · source rank 6the board, 29 ranked
  16. spreadsheet · Agents985 Elowin rate 20.5% · source rank 31the board, 32 ranked
  17. Web apps · Agents1181 Elowin rate 51.7% · source rank 27the board, 47 ranked
  18. 3D scenes · Models1225 Elowin rate 55.4% · source rank 42the board, 113 ranked
  19. ASCII art · Models1237 Elowin rate 58.8% · source rank 19the board, 67 ranked
  20. Code · Models1255 Elowin rate 54.0% · source rank 39the board, 120 ranked
  21. Data visualisation · Models1238 Elowin rate 53.6% · source rank 39the board, 110 ranked
  22. Game development · Models1253 Elowin rate 52.3% · source rank 36the board, 120 ranked
  23. SVG · Models1245 Elowin rate 58.0% · source rank 19the board, 83 ranked
  24. UI components · Models1264 Elowin rate 54.6% · source rank 36the board, 116 ranked
  25. Websites · Models1259 Elowin rate 53.8% · source rank 36the board, 126 ranked
Epoch Capabilities Index · a fit by a third party over every benchmark this model has results onthe whole index, with intervals
Index155.3153.5 to 157.3
Position45 of 258above 83% of the models placed
Results behind it18benchmark results
As Epoch names itGemini 3.5 Flashreleased 19 May 2026

Position 45 of the 258 models Epoch has placed, which is not the population of either board above, on a scale anchored at Claude 3.5 Sonnet 130 and GPT-5 150. Epoch states that absolute values are meaningless by themselves and that comparisons between models are what the index is for, that values are not linearly related to benchmark accuracy, and that a highly specialised model may score low while being capable in its own domain.

Benchmark results

2 of 6 boards · 45 results
Epoch AI Benchmarking Hub · publisher-run, CC-BY-4.015 of 22 tasks

Mean accuracy, 0 to 1, on the task the row names, as Epoch AI runs it and averaged over repeated runs. All eighteen of their tasks, including the held-out FrontierMath tiers, named as they name them. The model id carries the effort setting the run used. Epoch AI Benchmarking Hub names these runs gemini-3.5-flash_high, gemini-3.5-flash_low and gemini-3.5-flash_minimal, and the effort beside each score says which run it is. A run missing from a task's rows has no published result there, which is not a score of zero. All runs is how many runs the publisher scored on the task, every model's. Read from epoch.ai ↗, which dates its own data . Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.

LiveBench · publisher-run, Apache-2.023 of 23 tasks · 71 runs scored on each task · at high effort

Percent correct on the task the row names, higher is better. All 23 published columns; there is no average to carry, because the site computes its headline in the browser from a weighting this does not reimplement. LiveBench names this run gemini-3.5-flash-high. Read from livebench.ai ↗, which dates its own data . LiveBench, Apache-2.0. Results produced by the LiveBench maintainers on questions refreshed monthly.

2 of the 6 openly licensed boards Vitna mirrors carry a result for Gemini 3.5 Flash (batch), each printed as its publisher published it and none added to another. MTEB, HELM, Vectara hallucination leaderboard and Aider polyglot have not published one under a name that matches its listing, and names are matched to the catalog exactly and never guessed. A task that is a link opens its column on /capability, with every run scored on it; an effort that is a link opens its run's row there, where the run is ranked on Vitna's scale. Every board, with its licence and what its publisher measures, is on /benchmarks.

Sources

Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Source: the Epoch AI Capabilities Index (epoch.ai/eci), as of 20 Nov 2026. 258 models placed over 2917 benchmark results.
Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
Vitna ran none of these evaluations and adjusted none of these figures.