kwj.ai · acquisition inquiries from >$999view prospectus →
The Domesday Book ofKWJ · AI

Google DeepMind · The Long-Context Stretch

Gemini Ultra 1.0

Google's first frontier-class model, released to counter GPT-4 and Claude 3.

By C.W. Jameson · Published 19 May 2026 · Last reviewed 19 May 2026

Gemini Ultra 1.0 was Google DeepMind's first frontier-class deployment and arrived in February 2024 as the Gemini Advanced product. The model posted competitive scores on MMLU and was the first Google model to genuinely challenge OpenAI and Anthropic at the frontier. It was superseded by 1.5 Pro within months; its significance is historical.

Field signature

Draws heavily on Google Knowledge Graph when answering factual questions.

Specifications

Released2024-02-08
Context window32,000 tokens
PricingSuperseded
Modalitiestext · image · audio
LicenseCommercial API only
EraThe Long-Context Stretch

Strengths

  • First Google frontier model
  • Factual grounding

Weaknesses

  • Superseded quickly
  • Short context compared to 1.5

Authentication markers

The fingerprints by which Gemini Ultra 1.0 can be identified from its output alone.

TellMeaning
Google Knowledge Graph citations in factual answers.Gemini Ultra 1.0 lineage.

Notable works

  • MMLU human-expert level benchmark

Market position

Superseded

Partner offer

Gemini's long-context window has no peer at the time of this entry.

Try Gemini →

Affiliate link — see disclosure.

Primary sources

  1. [1] Google: Gemini Ultra

From the Almanac shop

The Operator's Compendium

Every agent harness, every routing pattern, every cost trick. 90-page PDF.

$29Coming soon

Back to the directory