AI Tool Guides, Comparisons & Latest News

Codex can show “thread not found” even when conversation history is still readable. Diagrams explain the difference between saved history and a conversation loaded for execution, followed by logs from a Windows case where sending recovered after reloading. Work through reopening, restarting, and checking a short response, then review read-only investigation and handoff options if the problem persists. Public counterexamples show why no universal fix can be claimed, and we have not verified a specific release that resolves the problem.

Jev is TypeSafe’s AI model for returning choices, scores, and Yes/No probabilities instead of generated prose. Learn how Choice, Score, and Noul differ, what confidence means, and how a support-routing API request is structured. Type guarantees do not guarantee correct judgments. This guide covers documented weaknesses involving calculations, dates, and misleading input, along with pricing, both input limits, and how to evaluate it before using it in languages other than English. It is based on official sources, not hands-on API performance testing.

When Claude Cowork was folded into chat on September 16, 2026, three creation features arrived in beta at the same time: Claude Docs for documents, Claude Slides for presentations and Claude Design for visual design. This article is about the first of them. In one sentence, Claude Docs turns what comes out of a conversation into a document you can keep editing. Ask it to write the discussion up as a spec the team can share, and Claude drafts it in front of you, asking about anything it is missing before it starts. What you get is rich text with headings and tables, and a single document can hold several tabs. You can edit it yourself, or select text inside the document, leave a comment and mention @Claude to have it make the change. The strength people overlook is that you can also turn a Claude Code session into a spec, a runbook or a report. It is a beta, though, and what is missing is very clearly missing: there is no version history, deletion cannot be undone, you cannot fix anything on mobile, and Team and Enterprise cannot share outside the organization. This article gives those absences as much space as the features, and works out what the tool is good for and what it is not.

Projects in Claude Code has been rebuilt. Until now a project was a folder holding conversations and reference material; the new Projects is a single conversation. You write what you need, Claude splits it into threads, the threads run in parallel in the cloud, and each one opens a pull request and reports back when it finishes. Closing your laptop does not stop them. Three things are worth checking before you jump in, though: the accounts that can use it are still limited (a Pro and Max public beta, going out first to accounts with no existing projects), github.com and the Claude GitHub App are requirements in practice, and the rate at which it eats your usage limit is nothing like a single session. This article works through how to tell whether the rollout has reached you, what a thread starts with (including the trap where permission rules and hooks stop applying once a project holds more than one repository), where the token cost comes from including the default of Opus at high effort, and how to choose between the five ways to work in parallel: subagents, agent view, agent teams, dynamic workflows and Projects, all from the official documentation and blog.

A newer model is smarter but more expensive, or so the thinking usually goes. Yet comparing GPT-6 Astra with GPT-5.6 Sol on unit price alone settles nothing, because Astra finishes the same job on fewer tokens. In measurements by the independent benchmarking firm Artificial Analysis, running Astra at its weakest setting, low, costs almost exactly what Sol costs at high for one task ($0.82 against $0.81) while scoring higher (46 against 42), on one third the output tokens and a quarter of the wait before the reply starts. Working from the figures as of September 18, 2026, this article covers cost, tokens and speed at each level of reasoning effort, the breakdown that explains why a 2.5x unit price still lands on the same total, what the numbers look like once Sol's promotional price ends, the way the gap in price narrows on coding agent work, and the steps for measuring all of it on your own work.

You want a smart model to handle only the planning, and a faster, cheaper model to handle the implementation. Claude Code's opusplan is a model setting that does this automatically. It runs Opus while in plan mode and Sonnet the rest of the time, and you can use it with /model opusplan or with model in settings.json. However, it does not appear in the /model list, and because the model switches every time you enter or leave plan mode, each switch re-reads the whole conversation without the cache. Based on the official documentation, the CHANGELOG and GitHub issues as of September 15, 2026, this article covers how to set it up (including pinning versions and 1M context), the flow from plan mode through approval to implementation, how it was removed from the selector in v2.0.0 and what an Anthropic staff member said about it, an estimate of the cache cost a switch creates and how to keep it down, how it differs from the advisor tool and subagents, and the kinds of work it suits and doesn't.

Can you keep Claude Code's main session on Opus 5 and hand only jobs like translation or high-volume checks to Sonnet or Haiku subagents? Yes. A subagent's model is decided in this order: the model passed at invocation, model in the definition file, the CLAUDE_CODE_SUBAGENT_MODEL environment variable, then the main session's model, and effort can also be set per subagent. Based on the official documentation as of September 15, 2026, this article covers how that order differs between versions, CLAUDE_CODE_SUBAGENT_MODEL_FORCE for pinning every subagent to one model, how aliases resolve differently by provider, and the fact that since v2.1.198 the built-in Explore inherits the main session's model. It then shares what happened when I actually launched subagents on other models and checked the conversation logs: they ran on the models I specified, each one read tens of thousands of tokens just to start, the subagent cache expires after 5 minutes even on a subscription, and the same translation given to Opus 5, Sonnet 5 and Haiku 4.5 twice each differed in time, cost and quality. Finally, it summarizes how this affects cost and usage limits, and what to weigh when deciding which work to move to a cheaper model.

Run several sessions in parallel and you start to wonder which one is eating your weekly limit. Yet Claude Code's /usage only shows the current session's numbers plus your plan-wide consumption split by skill, subagent, plugin and MCP server, and what share each session used appears neither in the desktop app's usage ring nor on the claude.ai settings page (as of September 2026). The answer is in the conversation logs stored on your machine (the JSONL files in ~/.claude/projects), but adding them up as they are gives the wrong result, because a single response is written across several lines, one per content block, and subagent records live in separate files. Measured on my own machine, the naive total came to about twice the correct value, and because the size of the error differed from session to session, even the ranking changed. This article covers what the official screens do and do not show, how to count the logs correctly with an aggregation script of about 50 lines, the measured result in which one session took nearly a third of all usage, the limits of what the numbers can tell you, and how to set up OpenTelemetry if you want to keep watching over time.

Right after you send a message, Gemini stops with "Something went wrong(13)", and Google does not explain the cause of this error in its official help. However, Google's official incident report from May 2026 records this message appearing for about four days, with the root cause given as insufficient database resources plus application bugs, and the workaround as "none". Building on that official record, and on the fact that the number "13" matches gRPC's internal error code (INTERNAL), though Google has not said they mean the same thing, this article sorts out the four situations that come up most in the official community: it hits everyone at once, only a long chat, only when an image is attached, and only one particular account. It then lays out seven isolation steps to try in order of least effort, from checking the status page to handing off to a new chat, testing a different account and reporting it through feedback.

You don't have to ask for "lead with the conclusion" or "keep it in plain English" every time: write it once in the instruction field that automatically applies to every conversation. That field is "Custom Instructions" in ChatGPT, "Instructions for Claude" in Claude and "Instructions for Gemini" in Gemini, but the three differ in what it is called, where it lives and how much you can write. ChatGPT allows 1,500 characters on Free and Go and 5,000 on Plus and higher plans (raised in July 2026), while Claude and Gemini don't publish a limit. This article checks the location and limit of each against the companies' official help, then uses Anthropic's official prompting guide to show how to write instructions that work, with an example. It also turns the situations the official help pages say instructions don't take effect, such as inside projects, inside Gems, in temporary chats and on company accounts, into a checklist.

You write to Claude in Spanish, Japanese or Portuguese, and the answer comes back in English. The official Claude Code repository keeps receiving the same report, and research has found that when the request and the answer are in different languages, even the strongest models fail to answer consistently in the language they were asked for. But there is more than one cause. Replies can drift into English gradually as Claude reads code and tool output, snap back to English right after compaction summarizes the conversation, or turn into a different language altogether, and each pattern responds to a different fix. This article shows how to tell the three apart, explains why the Claude Code language setting keeps working after compaction by pinning the instruction in the system prompt, and sorts out the problem reported from September 2026 of long sessions where the output itself breaks down, keeping confirmed information separate from what is still unconfirmed.

Install too many skills and they crowd out your context: half of that is right and half of it is wrong. According to the Claude Code documentation, the skill listing draws on a fixed budget of 1% of the model context window, and no matter how many skills you add, it stops there. Instead of growing, the listing starts dropping descriptions, beginning with the least frequently invoked skills, and keeps only their names. A skill that has lost its description no longer connects to what you asked for, and yet no error appears and nothing gets slower. This article works through the different jobs of the three measurement tools (/context, /usage and /skill-doctor), the definition of a cache miss as 5% and 2,000 tokens, how the cache lifetime falls from one hour to five minutes depending on your plan, why a CLI is still lighter than MCP now that tool definitions are lazy-loaded by default, the reasoning behind keeping CLAUDE.md under 200 lines, and what to cut first once you have measured, restricted throughout to what could be confirmed in the official documentation.