Claims and evidence
- Released Released as a preview (gpt-4-1106-preview) for all paying API developers; the stable model gpt-4-turbo-2024-04-09 followed on 2024-04-09. Primary source[1]GPT-4 Turbo with 128K context [2]Nov 6, 2023
- Deprecated Date of the notice covering gpt-4-turbo; the preview snapshots gpt-4-1106-preview and gpt-4-0125-preview were already deprecated in the notice of 2025-09-26. Primary source[3]2026-04-22: Legacy GPT model snapshots
- Updated preview gpt-4-0125-preview and the gpt-4-turbo-preview alias released Primary source[4]Updated GPT-4 Turbo preview [2]Jan 25, 2024
- GPT-4 Turbo with Vision generally available in the API (gpt-4-turbo-2024-04-09) Primary source[2]Apr 9, 2024 [5]Model details; Snapshots
- Status Deprecatedgpt-4-turbo (snapshot gpt-4-turbo-2024-04-09) is scheduled to shut down in the API on October 23, 2026. gpt-4-0125-preview and the gpt-4-turbo-preview alias were shut down on 2026-03-26 (notice of 2025-09-26); gpt-4-1106-preview is listed in both notices, with 2026-03-26 and October 23, 2026. Primary source[3]2026-04-22: Legacy GPT model snapshots
- Replaced by GPT-5.6 SolSubstitute model listed for gpt-4-turbo on the deprecations page when checked on 2026-10-01. Primary source[3]2026-04-22: Legacy GPT model snapshots
- Successor of GPT-4OpenAI calls GPT-4 Turbo the next generation of GPT-4. Primary source[1]GPT-4 Turbo with 128K context [5]description
- Change · Context length Context window of 128K tokens, which OpenAI says fits more than 300 pages of text; the gpt-4 API model had 8,192 tokens.Compared with GPT-4 Primary source[1] [6]
- Change · Training data Knowledge of world events extends to April 2023.Compared with GPT-4 Primary source[1]
- Change · Efficiency OpenAI priced it three times lower for input tokens and two times lower for output tokens than GPT-4.Compared with GPT-4 Primary source[1]
- Change · Availability Image input is part of the generally available API model gpt-4-turbo-2024-04-09 from April 2024, after a separate vision preview.Compared with GPT-4 Primary source[2] [1]
- Input text, imageImage input in the main model from general availability on 2024-04-09; the 2023 and January 2024 previews were text only, with images in the separate vision preview. Primary source[5]Model details [2]Apr 9, 2024
- Output text Primary source[5]Model details
- Feature Function calling Primary source[1]Function calling updates [5]Supported features
- Context window 128,000 tokensThe DevDay post gives 128K. Primary source[5]Model details [1]GPT-4 Turbo with 128K context
- Parameters Not disclosedNot disclosed by OpenAI. No claim madeNo source
- Access API Primary source[1]GPT-4 Turbo with 128K context [2]Apr 9, 2024
Lineage
Predecessors
- GPT-4 · 14 March 2023
Successors
- GPT-4o · 13 May 2024
Based on
Not derived from another model.
Variants and derived
None recorded.
Siblings
None recorded.
All ancestors
All descendants
- GPT-4o · 13 May 2024
- GPT-4.5 · 27 February 2025
- GPT-4.1 · 14 April 2025
- GPT-5 · 7 August 2025
- GPT-5.1 · 12 November 2025
- GPT-5.2 · 11 December 2025
- GPT-5.3 Instant · 3 March 2026
- GPT-5.4 · 5 March 2026
- GPT-5.4 mini · 17 March 2026
- GPT-5.5 · 23 April 2026
- GPT-5.5 Instant · 5 May 2026
- GPT-5.6 Luna · 9 July 2026
- GPT-5.6 Sol · 9 July 2026
- GPT-5.6 Terra · 9 July 2026
- GPT-5.6 Sol (August 2026 revision) · 6 August 2026
- GPT-6 Astra · 3 September 2026
- GPT-6 Luna · 22 September 2026
- GPT-6 Sol · 22 September 2026
- GPT-6.1 Sol · 29 September 2026
Variants
GPT-4 Turbo with vision
- Released Primary source[1]GPT-4 Turbo with vision
- Deprecated Primary source[3]2024-06-06: GPT-4-32K and Vision Preview models
- Retired Primary source[3]2024-06-06: GPT-4-32K and Vision Preview models
Inline variant in this record. Preview model for image input in the Chat Completions API (gpt-4-vision-preview). Vision moved into the main GPT-4 Turbo model at general availability on 2024-04-09, which the changelog also calls GPT-4 Turbo with Vision.
Status: Retired Evidence not assessed [3]2024-06-06: GPT-4-32K and Vision Preview models
Also known as: gpt-4-vision-preview, gpt-4-1106-vision-preview
Differs in:
- Input: text, image Evidence not assessed [1]GPT-4 Turbo with vision
Related AI Radar coverage
AI Radar coverage starts in June 2026; no coverage linked yet.