# Release Articoli collegati

Il Centro Notizie HTX fornisce gli articoli più recenti e le analisi più approfondite su "Release", coprendo tendenze di mercato, aggiornamenti sui progetti, sviluppi tecnologici e politiche normative nel settore crypto.

DeepSeek V4 'Full-Blooded Edition' Leaked, Could Be Released As Early As Tomorrow

The highly anticipated full release of DeepSeek V4 is imminent, expected to launch as early as tomorrow after nearly three months of waiting. A select group has already received access to the GA (General Availability) beta, which includes two versions: DeepSeek V4 Flash and DeepSeek V4 Pro. Early testers report that V4's overall performance is close to the level of Opus 4.8, with coding capabilities rivaling GPT-5.6 Sol. Its agent abilities are significantly enhanced, and 3D/SVG generation has improved notably. While it may not surpass the recently released Kimi K3 in performance, its expected price point is significantly lower. The official release will introduce a new "peak/off-peak" pricing model for its API. For example, deepseek-v4-pro will cost $0.87 per million output tokens during standard times and $1.74 during peak hours. The flash version is even more aggressive at $0.28/$0.56 per million tokens, with cached input tokens priced extremely low at $0.0028. This makes V4 a strong contender in terms of cost-effectiveness, potentially offering Opus-level capabilities at a fraction of the cost, continuing DeepSeek's reputation as a "price disruptor" in the AI market. Initial demos showcasing V4's capabilities have begun circulating, including generated 3D simulation games, HTML games blending elements of Minecraft and No Man's Sky, and classic games like a "Cut the Rope" clone. The final GA version is set to replace the older deepseek-chat and deepseek-reasoner models, which will be retired on July 24th.

marsbit07/19 05:31

DeepSeek V4 'Full-Blooded Edition' Leaked, Could Be Released As Early As Tomorrow

marsbit07/19 05:31

GPT-6 is Coming, OpenAI Completely Abandons the 4T Old Foundation

OpenAI is accelerating its roadmap, with GPT-6 rumored for imminent release and the complete abandonment of the older ~4 trillion parameter "Spud" base model. Meanwhile, the GPT-5.6 series has been officially greenlit by regulators and will launch publicly on Thursday, July 9th. This release features three specialized models: GPT-5.6 Sol (a high-performance flagship for complex coding, science, and security), GPT-5.6 Terra (a cost-effective generalist), and GPT-5.6 Luna (a high-speed, low-cost option for agents). Early testers praise Sol's relentless, "never-give-up" problem-solving approach, comparing its tenacity to a "Rottweiler" versus competitor Anthropic's Fable 5, described as a more insightful but sometimes haughty "wise owl." While some testers still favor Fable for certain creative or architectural tasks, GPT-5.6 Sol is highlighted for its dependable execution, strong sub-agent coordination, and significant improvements in coding efficiency, leading to a proposed optimal workflow: using Fable for brainstorming, Sol for execution, and Fable again for documentation.

marsbit07/09 05:08

GPT-6 is Coming, OpenAI Completely Abandons the 4T Old Foundation

marsbit07/09 05:08

Gemini 3.5 Pro Top Secret Leak, Front-end Capabilities Surpass Fable 5

"Gemini 3.5 Pro Leak: Frontend Surpasses Fable 5, Official Launch Expected July 17th. After a two-month silence, Google's Gemini 3.5 Pro appears poised for release. Leaks highlight its significant leap in frontend and visual code generation, reportedly surpassing competitor Fable 5. Key improvements include better design aesthetics, cleaner UI, enhanced SVG generation, and higher completion rates for full-page generation from single prompts. Tests show capabilities like creating detailed SVG portraits, complex UI elements, and 3D scenes with Three.js. However, the model reportedly still lags behind Fable 5 and GPT-5.6 in hardcore reasoning, agentic tasks, and long-range software engineering. The delay is attributed to Google reportedly conducting a new pre-training run with a fresh base model, rather than simply scaling up Gemini 3.5 Flash. Rumors also suggest this new base is being used for an image model, 'Nano Banana Pro,' targeting GPT-Image 2. The AI race intensifies as the industry awaits the potential July 17th reveal."

marsbit07/06 12:27

Gemini 3.5 Pro Top Secret Leak, Front-end Capabilities Surpass Fable 5

marsbit07/06 12:27

First Batch of GPT-5.6 Sol Internal Test Results Are In, Same-Task Cost is Only Half That of Fable 5

"GPT-5.6 Sol", the preview version of OpenAI's latest model, has released its first batch of internal testing results. According to an Nvidia engineer, Sol achieved CUDA acceleration results in 30 hours that took Opus 64 hours. The model is positioned for high-difficulty reasoning and complex, long-chain tasks like coding. Initial user feedback highlights that Sol produces more concise code—approximately 1/5 the lines of Opus—with a style closer to human-written code, favoring deep optimization over brute-force trial and error. In tests, Sol demonstrated superior performance in instruction following, spatial reasoning, and front-end design consistency compared to GPT-5.5 Pro. When compared to the recently relaunched competitor Fable 5, Sol shows slightly lower overall performance and code quality in some areas. For instance, Fable 5 can generate a playable game from a single prompt faster. However, Sol's key advantage is cost: at $5 per million input tokens and $30 per million output tokens, it is roughly half the price of Fable 5 ($10/$50). Additionally, users report Fable 5's safety restrictions are overly stringent, frequently blocking tasks, whereas Sol's safeguards are reportedly more tailored and less limiting. The model is expected to roll out fully to all users in the coming days.

marsbit07/06 07:29

First Batch of GPT-5.6 Sol Internal Test Results Are In, Same-Task Cost is Only Half That of Fable 5

marsbit07/06 07:29

Strike While the Iron is Hot: GPT-5.6's Three Major Models Fully Exposed, Scheduled for July 7th?

GPT-5.6 Models Leaked, Rumored for July 7th Release A leak from the Codex application's source code has revealed references to three upcoming GPT-5.6 sub-models: Sol, Terra, and Luna. A new "speed dial" feature was also discovered, suggesting users will be able to balance response speed and quality. According to rumors, OpenAI is targeting a release window of July 7th to 9th, strategically timed to coincide with the expiration of access limits for Anthropic's Claude Fable 5 model. Early comparative tests between a leaked GPT-5.6-terra version and Fable 5 show GPT-5.6 achieving similar or better results while using significantly fewer computational resources and responding faster. In complex coding tasks, GPT-5.6 demonstrated more efficient and user-collaborative problem-solving. The reported launch date appears calculated to capitalize on growing user frustration with competitor models. Anthropic's Fable 5 has faced criticism for overly aggressive safety filters that frequently downgrade responses and for severe "hallucination" issues in its Opus 4.8 model. Leaks also suggest GPT-5.6 Sol could be offered at a much lower price than Fable 5 while matching its performance. A final note warns developers with unused "rate limit reset" credits in Codex to use them before they expire around July 12th, ahead of the anticipated GPT-5.6 launch.

marsbit07/04 08:45

Strike While the Iron is Hot: GPT-5.6's Three Major Models Fully Exposed, Scheduled for July 7th?

marsbit07/04 08:45

Breaking News: The "Worker's Edition" Claude 5 Is Here, Everyone Can Use It

BREAKING: Claude Sonnet 5, dubbed "Fennec," is now the default model for all Free and Pro users. This mid-tier model boasts the strongest Agent capabilities in the Sonnet line yet, with performance rivaling the flagship Opus 4.8. It features autonomous planning and can utilize browser and terminal tools—capabilities previously exclusive to costly, large models. Key benchmarks highlight significant gains over its predecessor, Sonnet 4.6, in reasoning, tool use, coding, and knowledge work. Sonnet 5 scores 63.2% on SWE-bench Pro (surpassing GPT-5.5's 58.6%), 80.4% on Terminal-Bench 2.1, and 57.4% on Humanity's Last Exam (just 0.5% behind Opus 4.8). It even slightly outperforms Opus 4.8 in some knowledge tasks. Anthropic positions it as delivering ~90% of Opus's capability at a fraction of the cost. Pricing is aggressive: a limited-time promotional rate of $2 per million input tokens and $10 per million output tokens (reverting to $3/$15 after August 31). This undercuts Opus 4.8 ($5/$25) and GPT-5.5 ($5/$30). However, a new tokenizer may increase token counts by 1.0-1.35x, affecting final costs post-promotion. Notably, Sonnet 5 excels in security, with a mere 0.93% browser injection attack success rate, outperforming Mythos 5 and Opus 4.8. Its prompt injection defense matches Opus 4.8 at 0.19%. Launching amid uncertainty around the region-restricted Fable 5, Sonnet 5 is globally available. It targets the mid-market, offering near-flagship performance at a competitive price, effectively lowering the barrier for multi-Agent development and presenting a compelling alternative for cost-conscious developers.

marsbit07/01 07:47

Breaking News: The "Worker's Edition" Claude 5 Is Here, Everyone Can Use It

marsbit07/01 07:47

Just Now, Anthropic Released Sonnet 5, Performance Close to Opus 4.8, but Not Necessarily Cheaper

Anthropic has officially released Claude Sonnet 5, describing it as the most "agentic" Sonnet model to date. It can plan, use tools like browsers and terminals, and autonomously perform tasks at a level previously requiring larger, more expensive models. Performance in reasoning, tool use, programming, and knowledge work has significantly improved compared to Sonnet 4.6, now approaching that of Opus 4.8. Evaluation results indicate that Sonnet 5, at medium "effort" levels, offers better cost efficiency than its predecessor. At higher effort levels, its performance in some tasks can match Opus 4.8. In terms of safety, Sonnet 5 shows improved rates of refusing malicious requests and resisting prompt injection attacks compared to Sonnet 4.6, though it has a slightly higher rate of policy-violating behavior than Opus 4.8 and Mythos Preview. Its cybersecurity capabilities remain weaker than those models. Notably, Sonnet 5 uses a new tokenizer. The same text input now results in approximately 1.0 to 1.35 times more tokens, depending on content. To offset this, Anthropic offers a promotional launch price until August 31, 2026, at $2 per million input tokens and $10 per million output tokens. The standard pricing will be $3/$15 per million tokens thereafter. However, some external analysis suggests that due to increased token usage, the actual cost per task for Sonnet 5 may be higher than both Sonnet 4.6 and Opus 4.8.

marsbit07/01 00:35

Just Now, Anthropic Released Sonnet 5, Performance Close to Opus 4.8, but Not Necessarily Cheaper

marsbit07/01 00:35

GPT-5.6 Suffers 'Brain Surgery', Fable 5's Return to Be a Weaker Version?

GPT-5.6 and Fable 5 Face "Lobotomization" and Restrictions OpenAI has launched its new GPT-5.6 model in three limited tiers: Sol (advanced), Terra (balanced), and Luna (fast/economical). However, access to the most powerful Sol version is severely restricted, reportedly due to U.S. government intervention over safety concerns linked to its capabilities. Similarly, Anthropic's Fable 5 was globally disabled for 72 hours after a demonstration revealed its potential to exploit security vulnerabilities. It is now set to return in a heavily restricted "neutered" version for select partners only. Industry experts warn that such stringent safety "guardrails" could lead to "lobotomized" models with diminished intelligence and creativity. Developers are concerned about restricted access, potential extra costs, and models that no longer match their original promised capabilities. This marks a shift towards controlled access, contrasting with the earlier era of widely available, powerful AI.

marsbit06/29 08:37

GPT-5.6 Suffers 'Brain Surgery', Fable 5's Return to Be a Weaker Version?

marsbit06/29 08:37

Anthropic Released the "Most Powerful Model," But Most People Can't Use It

In April, Anthropic launched a preview of its "Mythos" model, which was not publicly released due to its exceptional ability to autonomously discover high-risk zero-day vulnerabilities, posing a security threat if misused. It was restricted to a trusted group of security partners under "Project Glasswing." On June 10, Anthropic officially released Fable 5 and Mythos 5. They share the same underlying model but are distributed under different rules. Fable 5 is for general users, while Mythos 5 remains locked for trusted security partners. Benchmarks show Fable 5 leading in software engineering and long-task execution, with significant improvements in generating production-ready code. However, Fable 5 includes a safety classifier that automatically downgrades requests related to cybersecurity, biochemistry, or model distillation to the weaker Opus 4.8 model. This mechanism, while intended for safety, can affect the user experience and has faced criticism for being overly conservative. Pricing is another key point. Fable 5's API costs are double that of Opus 4.8. Furthermore, after a free trial period ending June 23, it will be removed from standard subscription plans, requiring users to purchase additional credits for access. This shift signals a move towards pay-as-you-go pricing for the most advanced capabilities. The strategy highlights a growing divergence in the AI industry: while some players like DeepSeek are drastically cutting prices, Anthropic is increasing them for its top-tier model, using cost as a filter for high-value users. The article suggests the AI market is stratifying, with commoditized capabilities becoming cheaper while premium, cutting-edge models command a significant price premium.

marsbit06/10 23:52

Anthropic Released the "Most Powerful Model," But Most People Can't Use It

marsbit06/10 23:52

Just Now, Claude Mythos 5 Released, Handles 50 Million Lines of Code in One Day

Anthropic has unveiled Claude Fable 5 and the more powerful Claude Mythos 5, its new flagship AI models. Fable 5, available to most users, features a safety mechanism that downgrades risky queries to the older Claude Opus 4.8. The unrestricted Mythos 5 is limited to trusted users for sensitive fields like cybersecurity and bio-research. Key capabilities include exceptional performance in software engineering, completing a 50-million-line code migration in one day, and advanced visual reasoning, demonstrated by autonomously beating a Pokemon game using only screenshots. The models excel at long-context tasks, complex analysis in finance/law, and frontier scientific research. Mythos 5 has independently designed protein compounds now in drug development pipelines. Pricing is set at $10 per million input tokens and $50 per million output tokens. The release signifies a shift in human-AI interaction, with early testers noting it feels less like directing a tool and more like delegating to an autonomous "studio" capable of executing multi-hour projects from high-level instructions.

marsbit06/10 00:23

Just Now, Claude Mythos 5 Released, Handles 50 Million Lines of Code in One Day

marsbit06/10 00:23

活动图片

Technology Trendsright-arrow

Market Analysisright-arrow