ChatGPT: latest features and updates you should know about

ChatGPT has received major upgrades throughout 2025, making it more powerful, flexible, and accessible for developers, enterprises, and everyday users. With the release of GPT-5, the expansion of Advanced Voice Mode, new developer tools, and improved API capabilities, OpenAI has positioned ChatGPT as a comprehensive platform for productivity, automation, and multimodal interactions. This article explores the latest features and updates as of August-September 2025, based on verified and up-to-date information.

GPT-5 becomes available across plans.

The release of GPT-5 marks the most significant update to ChatGPT’s capabilities so far, delivering major improvements in reasoning, context handling, and multimodal processing.

• Launch date: 7 August 2025.

• Availability:

  • Free users: Limited GPT-5 access with a daily cap of around 20 messages.

  • ChatGPT Plus & Team: Unlimited GPT-5 availability, subject to soft throttling for heavy workloads.

  • Enterprise users: Full GPT-5 access, larger context windows, and higher throughput tiers.

• Pricing (API):

  • Base rates start at $1.25 per 1M input tokens and $10.00 per 1M output tokens.

  • Volume discounts apply for high-usage accounts, with rates dropping to $0.90 in / $7.20 out above 100M tokens per month.

• Key enhancements:

  • More accurate multi-step reasoning.

  • Faster response generation across web and API.

  • Optimized multimodal performance with improved image and document understanding.

This rollout positions GPT-5 as OpenAI’s new flagship model, offering stronger performance for developers and enterprise-grade workloads.

Advanced Voice Mode expands to all users.

ChatGPT’s Advanced Voice Mode is now available across all plans, delivering real-time natural speech interactions powered by improved text-to-speech engines.

• Rollout status: Completed on 17 July 2025 for mobile (iOS & Android) and 2 August 2025 for desktop web.

• Key improvements:

  • Over 50 supported languages for real-time voice conversations.

  • Lower first-token latency, making responses more immediate.

  • High-quality streaming voices optimized for long-form conversations.

• Deprecation of Standard Voice Mode:

  • Standard Voice Mode will be retired on 1 November 2025 in favor of Advanced Voice for better performance and wider accessibility.

This change makes voice-based conversations faster and more natural, aligning ChatGPT with real-time, assistant-style interactions.

ChatGPT Go introduces affordable access in India.

OpenAI launched ChatGPT Go, a region-specific subscription plan aimed at expanding accessibility in India.

• Launch date: 25 August 2025.

• Pricing: ₹399/month with UPI and Paytm billing options.

• Included benefits:

  • A dedicated 30,000 TPM API key for lightweight developer workloads.

  • Full access to Advanced Data Analysis (ADA) capabilities.

  • Access to GPT-4o and GPT-5 models without standard free-tier throttling.

• Education-focused initiative: OpenAI announced the India Learning Accelerator, providing 500,000 free ChatGPT licenses to schools and educators.

This launch reflects OpenAI’s broader strategy to localize pricing and improve adoption in high-growth regions.

The Responses API simplifies development workflows.

One of the most impactful updates for developers is the Responses API, which streamlines interaction patterns by merging multiple endpoints into a single, unified interface.

• Status: Currently in open beta since March 2025, with GA planned for later in 2025.

• New capabilities added in August 2025:

  • Binary file streaming alongside structured JSON responses.

  • Native multi-tool orchestration with parallel processing.

  • Support for server-sent event (SSE) streaming, improving performance in real-time apps.

• Use cases:

  • Automating document parsing pipelines.

  • Orchestrating agents that handle both tool calls and file inputs.

  • Creating single-call, multimodal workflows across chat, file search, and code execution.

Developers are encouraged to begin adopting the Responses API gradually, as it is set to replace /v1/chat/completions in late 2026.

Stricter JSON enforcement improves reliability.

In July 2025, OpenAI introduced support for strict JSON schema enforcement across GPT-4o, GPT-4.1, and GPT-5.

• Key update: Setting the parameter tool_definition.strict=true guarantees schema-perfect responses.

• Supported models: GPT-4o, GPT-4.1, GPT-5.

• Impact on developers:

  • Eliminates malformed JSON responses in structured outputs.

  • Reduces post-processing complexity when integrating with APIs.

  • Ensures predictable tool execution, especially in function-calling scenarios.

This update benefits developers working on production-grade integrations, where deterministic outputs are critical.

Larger context windows unlock enterprise-scale tasks.

OpenAI expanded context windows for enterprise accounts, improving performance for data-heavy applications and advanced analytics.

• Enterprise users: Up to 128,000 tokens of context for GPT-4o and GPT-5.

• Plus & Team plans: Retain a 64,000-token limit.

• Use cases enabled by larger context:

  • Processing massive PDFs, legal contracts, and technical manuals.

  • Building advanced knowledge retrieval systems.

  • Running multi-stage reasoning pipelines without context truncation.

This enhancement allows enterprise teams to process larger datasets and build AI agents capable of handling highly complex, multi-step workflows.

Gradients Workspace simplifies rapid prototyping.

OpenAI launched Gradients Workspace, a cloud-based IDE designed to make building ChatGPT-powered applications easier.

• Launch date: July 2025 (public preview).

• Key features:

  • Preconfigured GPT-4o kernel optimized for fast experimentation.

  • Integrated debugging tools and 1 GB of ephemeral storage per session.

  • Built-in code-interpreter sandbox supporting common data science libraries.

• Usage limits:

  • Sessions last up to 6 hours.

  • Resource caps: 4 vCPUs, 8 GB RAM.

Gradients Workspace helps developers prototype applications faster, without managing local environments or dependencies.

Other notable updates.

• India Learning Accelerator: Distributing 500,000 free licenses to schools as part of ChatGPT Go’s launch.

• Google Cloud integration: OpenAI added Google Cloud as a sub-processor in July 2025, improving performance and regional hosting options for Enterprise customers.

• Improved Advanced Data Analysis (ADA): Broader support for large file types and higher per-request size caps for Team and Enterprise tiers.

Comparing the latest ChatGPT features.

Feature	Launch / Status	Who benefits	Impact
GPT-5 Release	Aug 2025	All users	Advanced reasoning and multimodal understanding
Advanced Voice Mode	Jul 2025	Free, Plus, Team, Enterprise	High-quality, real-time speech support
ChatGPT Go Plan	Aug 2025	India-based users	Affordable GPT-5 and ADA access
Responses API	Beta (Mar 2025)	Developers	Unified endpoint for tools, chat, and file streaming
Strict JSON Mode	Jul 2025	API users	Schema-perfect structured outputs
Enterprise Context Expansion	May 2025	Enterprise users	Larger documents and more complex pipelines
Gradients Workspace	Jul 2025	Developers	Easier prototyping and deployment

Key takeaways.

• GPT-5 delivers major improvements in reasoning and multimodal performance, now available across all plans.

• Advanced Voice Mode now works for everyone, replacing legacy voice features.

• The Responses API introduces a simplified way to build multi-tool, multi-modal workflows.

• Developers benefit from strict JSON enforcement and cloud-based prototyping via Gradients Workspace.

• Enterprises gain 128k context windows, better compliance tools, and more robust hosting integrations.

With these updates, ChatGPT has evolved into a complete productivity platform, combining AI-powered reasoning, multimodal features, and enterprise-grade integrations in a unified ecosystem.

____________

FOLLOW US FOR MORE.

DATA STUDIOS

datastudios.org