Leaks
The latest AI Leaks from the internet.
DeepSeek V4.1 Flash Fast Spotted in Vercel AI Gateway
A report identifies DeepSeek V4.1 Flash Fast as a new variant and lists a 1.04858M context value plus text-and-image-to-text capability, although its release status remains unconfirmed.
Read the featureLeaks
-
Read →What One X Post Expects From OpenAI’s DevDay
An X post links OpenAI’s DevDay to possible plan changes, product releases and price cuts, but the supplied source does not confirm any announcement.
-
Read →Anthropic Fable 5.5: What the Unverified Report Says
An unverified X thread describes alleged grey-testing of Anthropic Fable 5.5, including unusually autonomous task completion and a possible launch…
-
Read →OpenAI Dots: What the Unverified Report Says
An unverified X report describes OpenAI Dots as a possible autonomous AI agent with its own virtual machine, approval controls, browser password…
-
Read →Claude Sonnet 5.5 Reportedly Appears in Claude Code Update
An X post reports that `claude-sonnet-5-5` appeared in Claude Code 2.1.284, but the evidence does not confirm a public launch or user access.
-
Read →Grok 4.8 Reportedly Appears in Cursor’s Model List
A reported Cursor server-side listing includes Grok 4.8 identifiers, but it does not confirm public access or a release date.
-
Read →Kimi K3.1 Appears in Backend Registry, but Release Is Unconfirmed
A report says the kimi-k3-1 slug survived a backend probe, but it does not confirm a public Kimi K3.1 release.
-
Read →Gemini 4 Pro Leak: Reported Benchmarks and Pricing
An unverified X thread and comparison image present alleged Gemini 4 Pro benchmark, context-window and pricing figures, but key details remain…
-
Read →ChatGPT’s Dots Reportedly Links Orbit and Aeon
An unverified report from a September 26 ChatGPT build links the dots label with orbit, aeon and related configuration and access-control identifiers.
-
Read →What the DeepSeek V5 Leak Actually Reports
An unverified X thread reports a possible DeepSeek V5 launch, rumored 2-trillion-parameter scale and Huawei Ascend training, but provides no official…
-
Read →Claude Sonnet 5.5 Specs Reportedly Listed in Droid
A Droid registry entry reportedly lists Claude Sonnet 5.5 with an 872,000-token context window, 128,000-token maximum output and several reasoning…
-
Read →Reported OpenCode AI Model Leak Names Five Models
An X post reports that OpenCode sitemap pages exposed five possible upcoming AI models, but the listing and any associated releases remain…
-
Read →Claude Sonnet 5.5 Reportedly Enters Claude Code Testing
A post reports grey testing, performance impressions and pricing for Claude Sonnet 5.5, but its availability and billing details remain unconfirmed.
-
Read →What OpenAI’s Reported Lowercase “o” Assistant Means
A reported ChatGPT code discovery appears to identify lowercase “o” as an assistant name across three Pro tier labels and 63 languages, but it does…
-
Read →What the MiniMax M3.1 Preview Report Says
A LuminaBench thread reports partner tests of a MiniMax M3.1 preview with selectable reasoning effort, image and video support, and DSpark…
-
Read →Dreamina Seedance 2.5 Premium Reportedly Adds 4K Video
A reported BytePlus model entry points to Dreamina Seedance 2.5 Premium with 4K output, up to 30-second video, and restricted access, but no official…
-
Read →Claude Sonnet 5.5 Release Date: What the Rumor Says
A post reports possible stealth testing of Sonnet 5.5 and suggests a near-term release, but provides no official confirmation or verifiable…
-
Read →Muse Spark 1.4 Listing Is a Signal, Not a Confirmed Launch
An OpenCode entry may relate to Meta’s teased next model, but the supplied evidence does not confirm Muse Spark 1.4 or reveal usable model details.
-
Read →OpenAI Codex Pro Max Support Reported in Backend Changes
A report points to ProMax support across Codex server-side systems, but the changes do not confirm public availability, pricing, quotas, or a launch.
-
Read →ChatGPT Pro Max: What the Report Shows
A post points to possible higher-tier ChatGPT Pro code references, but the screenshot does not confirm a product launch, final name, price or…
-
Read →Kling 4.0 leaked features: What the report says
An unverified X report describes possible Kling 4.0 video and image features, including 1080p output, longer videos and Omni Reference.
-
Read →Nano Banana 2.5 Flash Appears in Google Flow
A screenshot shared on X shows Nano Banana 2.5 Flash listed in a Google Flow model selector, but the model is not confirmed as available or…
-
Read →Gemini 4 Early Post-Training Release: What’s Reported
A report describes Google’s intention to release an early post-training version of Gemini 4, but the supplied evidence does not establish a firm…
-
Read →Claude Sonnet 5.5: What the Report Actually Says
A single X post reports that Anthropic is stealth testing Claude Sonnet 5.5, but the supplied evidence does not confirm a launch, public access or…
-
Read →What the Astra Minor Screenshot Actually Shows
A supplied screenshot appears to list Astra Minor as a standard mainline model with standard safeguards and security-related intended uses, but it…
-
Read →Claude Opus 5.5 Reported in Claude Code 2.1.280
An X post points to a screenshot listing claude-opus-5-5 in an apparent Claude Code 2.1.280 update, but release and availability remain unconfirmed.
-
Read →GPT-6-Sol Ultrafast Mode: What the Post Says
An X post speculates about a possible GPT-6-Sol Ultrafast mode and a speed of 750 tokens per second, but confirms neither detail.
-
Read →GPT-6 Sol, Luna and Astra: What the Azure Report Shows
An unverified X post points to Azure configuration entries for GPT-6 Sol, GPT-6 Luna and GPT-6 Astra Minor, but the supplied evidence does not…
-
Read →What Is OpenAI’s Rumored Aeon AI Assistant?
A post links the unconfirmed name Aeon to a possible OpenAI assistant, while an excerpt attributed to The Information describes related competition…
-
Read →What an X Post Says About GPT-6 Sol, Luna and Anthropic 5.5
A single, truncated X-post excerpt reports rumored OpenAI and Anthropic model updates, but does not independently confirm their launches,…
-
Read →Grok 4.7 Reportedly Spotted in OpenCode Zen
A truncated X excerpt reports that Grok 4.7 appeared in the OpenCode Zen harness, but it does not establish an official release, public access or…
-
Read →GPT-6 Sol API Support: What the Report Shows
A report from @LuminaBench points to possible GPT-6 Sol preparation across model and provider configurations, but does not confirm an official…
-
Read →Claude Opus 5.5 Reportedly in Stealth Testing
A post on X reports that Anthropic is testing Claude Opus 5.5 under the reported codename claude-wafer-eap, with a release planned for Tuesday. The…
-
Read →Opus 5.5 Pricing: Reported Token Rates
A supplied X post reports Opus 5.5 input and output rates, along with cache-read and cache-write figures whose billing units are unspecified.
-
Read →Anthropic Opus 5.5 Rumor: What the Report Shows
A brief post refers to Opus 5.5 in purported Anthropic internal code, but it does not establish an official model announcement or confirm skipped…
-
Read →Two Meta Model Names Reported in an X Post
An X post reports the names sam-3.1-contributor and sam-3.1, while its attached image provides no technical or official release details.
-
Read →What an X Post Reports About Gemini 4 Pro
A single X post attributes unusually large context and output limits, persistent memory, autonomous tools, malware handling and robot control to…
-
Read →Gemini 3.8 Live Reportedly Appears on Google Cloud Page
A post on X reports two Gemini 3.8 entries on a Google Cloud quota and metrics page, but it does not establish a launch, customer access or model…
-
Read →Claude Opus 5.2 May Be Tested in Claude Code
An X post suggests that a model related to Claude Opus 5.2 may be undergoing greyscale testing in Claude Code under the broader “Opus 5” name, while…
-
Read →What We Know About Google’s antigravity-preview-09-2026 Model
A post on X presents antigravity-preview-09-2026 as a new Google model, while suggesting a coding focus and updates to Antigravity and Jules agents…
-
Read →What Is GPT-6-sol? A Reported OpenAI Model Listing Explained
A reported screenshot shows the label “gpt-6-sol” alongside “Platform: OpenAI,” but it does not confirm an official announcement, release, or…
-
Read →GPT-6 “Sol, Terra and Luna”: What the Post Actually Says
A brief X post predicts future GPT model names, including GPT-6 Sol, Terra and Luna and GPT-5.6 Luna, but provides no release details or confirmation.
-
Read →GPT-6 Leak? Astra Slugs Appear in Codex Desktop
Two GPT-6-related slugs reportedly appeared in a Statsig feature flag in Codex Desktop, prompting speculation about a forthcoming model or product…
-
Read →ImageGen 2.5 Leak Hints at a Major Image Upgrade
A leaked Codex screenshot appears to reveal an ImageGen 2.5 announcement promising higher-quality results, faster generation, and smarter creative…
-
Read →GPT-6 Astra Release Theory Emerges From API Probe
An API probe shared on X has sparked speculation that GPT-6 Astra may be staged behind OpenAI’s infrastructure, but the evidence remains unverified.
-
Read →GPT-6 Name Report Raises Questions About Astra
A report cited in an X post suggests OpenAI considered, then rejected, the GPT-6 name for Astra. That leaves GPT-5.7 Astra and GPT-Astra as possible…
-
Read →OpenAI Astra Leak Points to New Model Testing
An unconfirmed report says OpenAI may be testing model variants associated with the rumored Astra project, including a new vega-alpha build.
-
Read →Grok 4.7 Release Set for 10 Days After Announcement
A post from @elonmusk says Grok 4.7 is expected to come out in 10 days, suggesting an estimated release around September 12, 2026.
-
Read →Claude Code 2.1.257 Adds Fable 5.1 and Mythos 5.1
An X post reports that the imminent Claude Code 2.1.257 release adds Claude Fable 5.1 and Mythos 5.1 to the model catalog, with expanded context and…
-
Read →Fable 5.1 Appears Staged on Amazon Bedrock
A change in Amazon Bedrock’s API response suggests that Fable 5.1 may be approaching release, though the model is not confirmed as publicly available.
-
Read →Astra Reportedly Moves Into Partner Early Access
A report says Astra has moved beyond dogfood testing under the name “ultima-alpha,” with broader access and a GPT-Image 2 update potentially arriving…
-
Read →GPT Astra Internal Testing Reportedly Expands
A post reports expanded internal testing of OpenAI’s GPT Astra, early public outputs and a possible frontend improvement, while no public release has…
-
Read →Fable 5.1 has been spotted..
Fable 5.1 could be live for you now
No entries match that filter.