For direct Anthropic API access, set ANTHROPIC_API_KEY and approve it in the interactive CLI. Gateways need their own endpoint and credential; cloud providers and Desktop use separate configuration paths.
Sonnet 5.5 keeps Sonnet 5's $2/$10 token prices. On Artificial Analysis tests it costs less per task from low to xhigh effort and about 51% more at max.
Temporarily unable to authenticate comes from Anthropic: stay signed in and retry. Login expired or Failed to authenticate means /login. The wording decides.
A Codex token-exchange 403 is a refused authentication request. The full error tells you whether to repair callback access, check the client’s network, renew stored credentials, or stop for an access restriction.
GPT-6.1 Sol keeps GPT-6 Sol's $2/$10 rates but halves cached input to $0.10. Requests using none effort or Chat Completions tools must change.
For APIConnectionError, capture the underlying exception and test DNS, TLS, proxy settings, and the exact endpoint from the failing runtime. An HTTP response narrows the problem; recovery still requires a successful request through the original application.
For OpenAI API quota exceeded or insufficient_quota, inspect error.code before buying credits or retrying. A depleted balance, enforced spend cap, and approved usage limit need different fixes; ChatGPT subscriptions do not fund API calls.
Compare the best AI video models in 2026 including Sora 2, Veo 3.1, Runway Gen-4.5, Kling 2.6, and open-source options. Features verified pricing data, real cost calculations per 10-second video, scene-based recommendations, and an API developer guide.
A returned Claude API 500 api_error calls for bounded retries and request evidence. If it keeps failing, stop the immediate loop, defer safe work, and send a redacted reproduction to the service owner.
Claude Code API Error 500 is a server-side failure, not your prompt or account. Check status, then retry, sign in again, or resume after checking finished work.
A slash command is an invocation interface; a skill packages reusable instructions; a hook runs at a lifecycle event. Custom commands now use the skills system, and /claude-api invokes the bundled Claude API skill.
"Rate limit reached" in Claude Code is usually a session, weekly, or Opus/Sonnet limit. Match the line, then resume in the same session without resending.
Run /usage for Claude Code's 5-hour session and weekly limits and their reset times. /status only shows account and model. Audit fast drain before upgrading.
Claude API Error: Rate limit reached and Rate Exceeded are not one limit. Match the current message to plan usage, 1M access, API key, provider, or gateway.
Gemini 3 Pro Image (Nano Banana Pro) has no free API tier and no free Pro quota in the Gemini app. The one no-cost route is AI Mode in Search for US adults.
Gemini API rate limits (RPM, TPM, RPD) apply per project, not per key, and your live numbers are in AI Studio. Tier 2 and Tier 3 unlock at $100 and $1,000 paid.
A Gemini image 429 is a per-minute limit (clears in a minute), a daily quota (resets midnight Pacific), a spend cap per 10 minutes, or a free-tier limit of 0.
No. As of October 5, 2026, no Gemini image model has free API output. The $5 minimum prepay covers about 148 Nano Banana 2 Lite images at 1K.
As of October 5, 2026, the Gemini image API runs on three models, from $0.0336 per image. Image output has no free tier: you need a paid key with $5 prepaid.
Upload the photo to Gemini, ask it to change only the background, then check the face and edges. Gemini can't make a transparent PNG; use a cutout tool.
Gemini app image failures are usually sign-in, age, quota or a policy refusal. In the API, retry 429 and 5xx with backoff, and never retry 400, 402 or 403.
Gemini creates and edits people images within consent, privacy and safety rules. A no-person image request helps separate a people refusal from an access or quota problem.
Gemini API Tier 3 (T3) isn't a special key: the billing account upgrades itself after $1,000 in paid Google Cloud spend and 30 days from the first payment.
Seedance 2.0 is ByteDance's AI video generation model that creates up to 15-second clips at 2K resolution with native audio. This guide covers 4 access methods (including free Dreamina), step-by-step tutorials for text-to-video, image-to-video, the 12-file multimodal reference system, lip sync in 8+ languages, commercial use guidance, and current limitations with workarounds.