{ "guide": { "slug": "gpt-5-3-codex-spark-vs-gpt-5-5-sonnet", "title": "GPT-5.3-Codex-Spark vs GPT-5.5 vs Claude Sonnet Comparison Guide", "subtitle": "A practical routing guide for choosing the fast Codex model, the frontier OpenAI model, or Anthropic's current Sonnet tier for coding-agent work.", "summary": "Use GPT-5.3-Codex-Spark for latency-bound interactive coding loops, GPT-5.5 for quality-bound Codex and professional work, and Claude Sonnet 4.6 when Anthropic's speed/intelligence balance, 1M context, or Claude Code surface is the better fit.", "heroLabel": "Model Routing", "vaultPath": "Research/Technology/Software Development/AI Tools/GPT-5.3-Codex-Spark vs GPT-5.5 vs Claude Sonnet Comparison Guide.md", "version": "2026-05-09.1", "publishedAt": "2026-05-09T00:00:00-04:00", "updatedAt": "2026-05-09T00:00:00-04:00", "r2JsonKey": "guides/gpt-5-3-codex-spark-vs-gpt-5-5-sonnet/2026-05-09.1.json", "category": { "slug": "software", "name": "Software", "description": "Fact sheets for developer tools, AI systems, software platforms, and technical buying or adoption decisions." }, "tags": [ "ai-tools", "coding-agents", "model-routing", "openai", "anthropic" ], "freshness": "Verified against OpenAI and Anthropic public model pages, API docs, launch posts, and system-card material on 2026-05-09. Re-check access, pricing, context limits, and benchmark tables before budgeting automated usage because these model surfaces move quickly.", "stats": [ { "label": "Spark context", "value": "128k", "detail": "Text-only research preview at launch." }, { "label": "Spark speed", "value": "1000+ tok/s", "detail": "OpenAI claim for ultra-low-latency Cerebras serving." }, { "label": "GPT-5.5 Codex context", "value": "400k", "detail": "Codex context window in OpenAI's checked launch note." }, { "label": "GPT-5.5 API price", "value": "$5 / $30", "detail": "Per 1M input/output token pricing in OpenAI's checked API model page." }, { "label": "GPT-5.5 API context", "value": "1.05M", "detail": "OpenAI model page lists 1,050,000 context and 128,000 max output tokens." }, { "label": "Sonnet context", "value": "1M", "detail": "Claude Sonnet 4.6 context in Anthropic's model table." }, { "label": "Sonnet price", "value": "$3 / $15", "detail": "Per 1M input/output token pricing in Anthropic's model table." } ], "thesisCards": [ { "title": "Spark is for the tight loop.", "body": "Use GPT-5.3-Codex-Spark when the bottleneck is waiting for the model, not deciding what the patch should be." }, { "title": "GPT-5.5 is for ownership.", "body": "Use GPT-5.5 when the agent needs to inspect, reason, edit, run checks, and take responsibility for a harder task." }, { "title": "Sonnet is the Claude workhorse.", "body": "Use Claude Sonnet 4.6 when the job is in Claude Code or when Anthropic's 1M-context speed/intelligence tradeoff is the better operating surface." } ], "sections": [ { "id": "bottom-line", "eyebrow": "Bottom line", "title": "Route by bottleneck, not by benchmark headline.", "body": "Spark is the fast interactive Codex model. GPT-5.5 is the stronger OpenAI model to reach for when correctness, repo-wide reasoning, or autonomous execution matters. Claude Sonnet 4.6 is Anthropic's current Sonnet tier and a strong daily-driver choice when Claude Code, Anthropic APIs, 1M context, or adaptive thinking fit the job." }, { "id": "snapshot", "eyebrow": "Snapshot", "title": "The useful stats", "table": { "columns": [ "Axis", "GPT-5.3-Codex-Spark", "GPT-5.5", "Claude Sonnet 4.6" ], "rows": [ [ "Best role", "Real-time Codex collaboration", "Frontier Codex and professional work", "Fast, capable Claude agent backbone" ], [ "Public status", "Research preview", "Public API model page plus ChatGPT and Codex rollout", "Current Sonnet model" ], [ "Context", "128k", "400k in Codex; 1,050,000 in API", "1M" ], [ "Inputs", "Text-only at launch", "Text and image input; text output", "Text and image input" ], [ "Speed signal", "More than 1000 tokens/sec on ultra-low-latency Cerebras serving", "Fast mode in Codex is 1.5x faster for 2.5x cost", "Fast comparative latency in Claude docs" ], [ "Pricing signal", "No general API price found; separate preview limits", "$5/M input, $0.50/M cached input, $30/M output; long prompts over 272k input tokens cost more", "$3/M input, $15/M output" ], [ "Default behavior", "Lightweight, targeted; does not auto-run tests unless asked", "API reasoning effort supports none, low, medium default, high, and xhigh", "Fine-grained thinking effort, extended thinking, and adaptive thinking" ] ] } }, { "id": "routing-rules", "eyebrow": "Decision ladder", "title": "When to reach for each model", "table": { "columns": [ "Situation", "First reach", "Why" ], "rows": [ [ "You are watching the agent and want fast iteration", "GPT-5.3-Codex-Spark", "The model is designed for real-time coding, interrupts, redirects, and rapid steering." ], [ "You need one small patch in a familiar file", "GPT-5.3-Codex-Spark", "Its lightweight default is a feature when the desired diff is narrow." ], [ "You need a fast review or scout pass before asking a bigger model", "GPT-5.3-Codex-Spark", "Low latency makes it useful for second opinions and quick triage." ], [ "You need the agent to inspect a large repo, reason through tradeoffs, run tests, and land a change", "GPT-5.5", "The task is quality-bound and tool-bound rather than latency-bound." ], [ "You need a hard architecture call, difficult bug, or high-confidence review", "GPT-5.5", "OpenAI positions it as the stronger professional and Codex tier." ], [ "You are in Claude Code or building against Anthropic APIs", "Claude Sonnet 4.6", "It is Anthropic's current speed/intelligence workhorse with 1M context." ], [ "You want the strongest Anthropic model for deep refactors or multi-agent coordination", "Claude Opus, not Sonnet", "Anthropic says Opus remains the deepest-reasoning choice for those jobs." ] ] } }, { "id": "benchmark-notes", "eyebrow": "Benchmarks", "title": "Stats are directional, not a clean scoreboard.", "table": { "columns": [ "Metric family", "GPT-5.3-Codex-Spark", "GPT-5.5", "Claude Sonnet 4.6" ], "rows": [ [ "SWE-style coding", "OpenAI says strong SWE-Bench Pro performance in a fraction of GPT-5.3-Codex time, but the public page did not expose a numeric score in text.", "58.6% on SWE-Bench Pro Public in OpenAI's table.", "Anthropic system card reports 79.6% SWE-bench Verified; this is not the same benchmark." ], [ "Terminal / agentic coding", "OpenAI says strong Terminal-Bench 2.0 performance faster than GPT-5.3-Codex, no numeric text found.", "82.7% on Terminal-Bench 2.0.", "Anthropic reported scores vary by harness and thinking setup; compare carefully." ], [ "Computer use", "Not the right target; Spark is text-only at launch.", "78.7% OSWorld-Verified in OpenAI's table.", "Anthropic positions Sonnet 4.6 as much improved for computer use." ], [ "Tool use", "Useful in Codex's real-time loop, but public source emphasizes latency more than tool benchmark scores.", "75.3% MCP Atlas and 55.6% Toolathlon in OpenAI's table.", "Strong tool-selection and error-correction positioning; use local evals for direct comparison." ] ] } }, { "id": "operating-playbook", "eyebrow": "Playbook", "title": "A simple routing policy", "bullets": [ "Start with Spark when the task should fit in one tight interactive loop.", "Escalate to GPT-5.5 when Spark starts asking for broad context, proposes a risky diff, or needs to own verification.", "Use Sonnet 4.6 when Claude Code is already the best product surface, when 1M-context Sonnet economics matter, or when local evals show stronger instruction following on your codebase.", "Route critical refactors and production-risk reviews to the most capable model available, then use a faster model for follow-up cleanup after the hard part is settled.", "Keep local scorecards. Vendor stats choose the first model; your repo's failure modes decide the durable default." ] }, { "id": "caveats", "eyebrow": "Caveats", "title": "What can go stale", "bullets": [ "Spark access and limits are research-preview facts.", "GPT-5.5 now has a public API model page, but long-context pricing and tool-call fees can materially change automation cost.", "Sonnet means Sonnet 4.6 in this guide because that was the current Anthropic Sonnet page and model table on 2026-05-09.", "Do not compare Sonnet 4.6 SWE-bench Verified directly to GPT-5.5 SWE-Bench Pro Public. They are different benchmarks.", "Public pages do not capture every product-plan limit, enterprise admin setting, or Codex/Claude Code UI behavior." ] } ], "sources": [ { "id": "openai-spark-launch", "title": "Introducing GPT-5.3-Codex-Spark", "publisher": "OpenAI", "url": "https://openai.com/index/introducing-gpt-5-3-codex-spark/", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "openai-gpt55-launch", "title": "Introducing GPT-5.5", "publisher": "OpenAI", "url": "https://openai.com/index/introducing-gpt-5-5/", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "openai-gpt55-model", "title": "GPT-5.5 model page", "publisher": "OpenAI API docs", "url": "https://platform.openai.com/docs/models/gpt-5.5", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "anthropic-sonnet-page", "title": "Claude Sonnet 4.6", "publisher": "Anthropic", "url": "https://www.anthropic.com/claude/sonnet", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "anthropic-sonnet-launch", "title": "Introducing Claude Sonnet 4.6", "publisher": "Anthropic", "url": "https://www.anthropic.com/news/claude-sonnet-4-6", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "anthropic-models", "title": "Models overview", "publisher": "Anthropic API docs", "url": "https://platform.claude.com/docs/en/about-claude/models/overview", "accessedAt": "2026-05-09", "sourceType": "official" }, { "id": "anthropic-system-card", "title": "Claude Sonnet 4.6 System Card", "publisher": "Anthropic", "url": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "accessedAt": "2026-05-09", "sourceType": "official" } ], "claims": [ { "id": "spark-positioning", "claim": "GPT-5.3-Codex-Spark is a research-preview, smaller Codex model designed for real-time coding in Codex.", "sourceId": "openai-spark-launch", "evidenceType": "official", "confidence": "high", "caveat": "Research-preview access and limits may change quickly.", "guideLocation": "Bottom line" }, { "id": "spark-context", "claim": "Spark is text-only with a 128k context window at launch.", "sourceId": "openai-spark-launch", "evidenceType": "official", "confidence": "high", "caveat": "OpenAI says later ultra-fast models may add larger context and multimodal input.", "guideLocation": "Snapshot" }, { "id": "spark-speed", "claim": "Spark is optimized for latency and can deliver more than 1000 tokens per second on ultra-low-latency Cerebras serving.", "sourceId": "openai-spark-launch", "evidenceType": "official", "confidence": "high", "caveat": "The speed claim is hardware and serving-path dependent.", "guideLocation": "Stats" }, { "id": "spark-default-style", "claim": "Spark keeps a lightweight default working style and does not automatically run tests unless asked.", "sourceId": "openai-spark-launch", "evidenceType": "official", "confidence": "medium", "caveat": "This is release-note behavior, not a guaranteed invariant across all Codex surfaces.", "guideLocation": "Snapshot" }, { "id": "gpt55-availability", "claim": "GPT-5.5 is available in ChatGPT and Codex, with Codex availability across Plus, Pro, Business, Enterprise, Edu, and Go plans and a 400k Codex context window.", "sourceId": "openai-gpt55-launch", "evidenceType": "official", "confidence": "high", "caveat": "Subscription and admin settings may affect access.", "guideLocation": "Snapshot" }, { "id": "gpt55-price-context", "claim": "GPT-5.5 is listed on the official API model page at $5/M input, $0.50/M cached input, and $30/M output, with a 1,050,000-token context window, 128,000 max output tokens, text and image input, text output, and reasoning effort values of none, low, medium default, high, and xhigh.", "sourceId": "openai-gpt55-model", "evidenceType": "official", "confidence": "high", "caveat": "Long prompts over 272k input tokens have higher pricing; check the current pricing page before large-context automation.", "guideLocation": "Stats" }, { "id": "gpt55-coding-benchmarks", "claim": "GPT-5.5 scores 58.6% on SWE-Bench Pro Public and 82.7% on Terminal-Bench 2.0 in OpenAI's published table.", "sourceId": "openai-gpt55-launch", "evidenceType": "official", "confidence": "medium", "caveat": "Vendor benchmark tables are not harness-normalized against Spark or Sonnet.", "guideLocation": "Benchmark notes" }, { "id": "gpt55-tool-computer-benchmarks", "claim": "GPT-5.5 scores 78.7% on OSWorld-Verified and 75.3% on MCP Atlas in OpenAI's published table.", "sourceId": "openai-gpt55-launch", "evidenceType": "official", "confidence": "medium", "caveat": "Computer-use and tool-use results depend heavily on scaffold and allowed tools.", "guideLocation": "Benchmark notes" }, { "id": "sonnet-current", "claim": "Current Claude Sonnet means Claude Sonnet 4.6 as of 2026-05-09.", "sourceId": "anthropic-models", "evidenceType": "official", "confidence": "high", "caveat": "Anthropic may ship a newer Sonnet later; re-check before relying on this guide.", "guideLocation": "Snapshot" }, { "id": "sonnet-stats", "claim": "Claude Sonnet 4.6 is positioned as the best combination of speed and intelligence, with 1M context, 64k max output, fast comparative latency, extended thinking, adaptive thinking, and $3/M input plus $15/M output pricing.", "sourceId": "anthropic-models", "evidenceType": "official", "confidence": "high", "caveat": "Third-party platform routing and pricing may vary.", "guideLocation": "Snapshot" }, { "id": "sonnet-agent-positioning", "claim": "Anthropic positions Sonnet 4.6 for coding, agents, computer use, tool selection, error correction, and production workflows.", "sourceId": "anthropic-sonnet-page", "evidenceType": "official", "confidence": "high", "caveat": "Positioning is vendor-supplied.", "guideLocation": "Routing rules" }, { "id": "opus-caveat", "claim": "Anthropic still recommends Opus for tasks demanding the deepest reasoning, including codebase refactoring and multi-agent coordination where precision is paramount.", "sourceId": "anthropic-sonnet-launch", "evidenceType": "official", "confidence": "high", "caveat": "This comparison is within Anthropic's own model family, not against GPT-5.5.", "guideLocation": "Routing rules" }, { "id": "benchmark-comparability", "claim": "Direct GPT-5.3-Codex-Spark vs GPT-5.5 vs Sonnet benchmark comparisons should be treated cautiously unless harness, effort setting, tools, sample count, prompt, and budget are normalized.", "sourceId": "openai-gpt55-launch", "evidenceType": "synthesis", "confidence": "high", "caveat": "This is an inference from vendor benchmark heterogeneity, not a single published benchmark result.", "guideLocation": "Benchmark notes" } ], "revisions": [ { "version": "2026-05-09.1", "date": "2026-05-09", "changes": [ "Created the initial comparison guide for GPT-5.3-Codex-Spark, GPT-5.5, and Claude Sonnet 4.6.", "Added model-routing rules, stats, benchmark caveats, source rows, and claim receipts.", "Published as a public Field Guides page with Software taxonomy placement." ] } ] }, "publicUrl": "https://fieldguides.co/g/gpt-5-3-codex-spark-vs-gpt-5-5-sonnet" }