Historical identifier; current API calls automatically route to gemini-3.8-flash.
Access
Original release is deprecated and is no longer selectable in Google AI Studio. The identifier remains accepted, but routes to gemini-3.8-flash. Original-checkpoint access through other channels is not established.
Regions
Not collected for the original release
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 14 Oct 2026
Category
Usage
Rate
Unit
Historical Google pricing announcement
Original introductory input1
$0.75
1M tokens
Historical Google pricing announcement
Original introductory output1
$3.75
1M tokens
Historical Google pricing announcement
Previously announced January 2027 input1
$1.50
1M tokens
Historical Google pricing announcement
Previously announced January 2027 output1
$7.50
1M tokens
Rate conditions
August 2026 model card: introductory rates scheduled through December 31, 2026, with higher rates announced for January 1, 2027. The original model selector now redirects. This preserved announcement is not a current original-checkpoint quote or proof the future schedule remains applicable. Cache and additional charges are not established here.
Current host-listed selector. Underlying served checkpoint after Google migration is not verified.
Access
Perplexity API key; POST /v1/agent, default service tier. The current host catalog retains this selector and price; original-checkpoint availability is not established.
Current Perplexity host-listed default-tier rates for this legacy selector. Original weights and host-specific token limits after migration are not verified. Tool invocations may add charges. Not Google direct prices. Cached input price: $0.075 / 1M tokens (base tier).
Current Perplexity host-listed default-tier rates for this legacy selector. Original weights and host-specific token limits after migration are not verified. Tool invocations may add charges. Not Google direct prices.
Observed 14 Oct 2026. Write rates and storage fees are different charges; use the billing rules above.
Caching reviewed 14 Oct 2026
What the measurements say
Artificial Analysis Intelligence Index
Google DeepMind · provider-reported
56index points
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Composite index, not percent accuracy. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Google cites official public leaderboard. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed, mini SWE agent with LiteLLM 1.96 and high thinking. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Google cites public Web Dev leaderboard. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed with default Terminus-2 harness. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed mini SWE agent with LiteLLM 1.96; cli-2ph-simplex and vpp-loss-divergence modified for container compatibility. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Private set, result attributed by Google to the public leaderboard. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Google cites Artificial Analysis public leaderboard. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Google cites Artificial Analysis public leaderboard. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed expert PDF comprehension evaluation. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed without tools. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed with search and code execution. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed, 1,024 frames, no tools. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Cumulative 128K score, self-computed. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Partial score; best of three runs, one attempt/run, 1080p, up to 500 steps, screenshot observations and pyautogui. Before the August 8 benchmark patch. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed with default ALE-Claw harness. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed on 1,811 verified items: 668 original plus 1,143 revised. Excludes 689 uncertain original items; not original HLE. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Linux terminal, bioinformatics tools, Python/R and benchmark-allowlisted internet domains. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Separate difficult subset with the same terminal and restricted internet setup. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Original August 2026 evaluation, as reported by Google; not independently reproduced here. Self-computed with Linux terminal, bioinformatics tools, Python/R and internet access. Methodology: https://deepmind.google/models/evals-methodology/gemini-3-7-flash . Default API sampling and pass@1 unless the specific methodology says otherwise.
Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.
The record behind the model
Why it is included
Google documents a distinct original Gemini release with an identified base, launch history, model card and dated evaluations. Distribution is through managed products and APIs; no public weights are offered in these reviewed cards. This is a distribution inference, not an explicit weight license.
