Compare Claude Fable 5 and Claude Sonnet 5: Expressive Mastery vs Enterprise Agents
Claude Fable 5 vs Claude Sonnet 5: which should I use?
Claude Sonnet 5 is Anthropic’s flagship for coding and agents ($2.00/$10.00 per million tokens, 200K context), while Claude Fable 5 is built for complex narrative synthesis and 1M context ($10.00/$50.00 per million tokens). Verified 2026-09-07.
Verified 2026-09-07
Where can you find price, speed, and task evidence for Claude Fable 5 and Claude Sonnet 5?
Which tasks fit Claude Fable 5 and Claude Sonnet 5?
Best-fit task signals from the published model strengths; this is not a substitute for a controlled benchmark.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| Best fit | Legacy ultra-tier Claude workloads superseded by Fable 5.1. | Production workloads that need near-Opus quality with better speed and cost. |
| Reasoning mode | Available | Available |
What does a Coding Agent workload cost with Claude Fable 5 and Claude Sonnet 5?
Modeled for Coding Agent: 20,000 input + 2,000 output tokens per task, 1 turn. This is a workload model, not a provider quote.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| Coding Agent / task | $0.30 (modeled) | $0.06 (modeled) (winner) |
| Input / output rate | $10.00 / $50.00 per M | $2.00 / $10.00 per M |
How fast are Claude Fable 5 and Claude Sonnet 5?
Only non-estimated benchmark results are shown as measured.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| Measured throughput | 41 tokens/s | Unavailable |
| Time to first token | 610 ms | Unavailable |
How compatible are Claude Fable 5 and Claude Sonnet 5 with APIs?
Provider-level wire-format and SDK facts; model-specific parameter differences may still apply.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| OpenAI SDK | Not drop-in | Not drop-in |
| Request shape | Messages API — system prompt is a top-level field, not a message role, and max_tokens is required on every call. | Messages API — system prompt is a top-level field, not a message role, and max_tokens is required on every call. |
| Streaming | SSE (Anthropic content-block events) | SSE (Anthropic content-block events) |
What are the privacy and retention policies for Claude Fable 5 and Claude Sonnet 5?
Provider policy and residency facts from the verified provider registry. A missing retention policy is not treated as a guarantee.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| Provider says API data trains models | No | No |
| Published retention period | Unavailable | Unavailable |
| Data residency | Unavailable | Unavailable |
How much effort does it take to migrate between Claude Fable 5 and Claude Sonnet 5?
Direction is from each displayed model to the other. Effort is derived from provider compatibility and published parameter maps.
| Fact | Claude Fable 5 | Claude Sonnet 5 |
|---|---|---|
| Move Claude Fable 5 → Claude Sonnet 5 | drop-in; 0 breaking parameter differences | Target: Claude Sonnet 5 |
| Move Claude Sonnet 5 → Claude Fable 5 | Source: Claude Sonnet 5 | drop-in; 0 breaking parameter differences |
| Why | Same provider; only the model identifier changes. | Same provider; only the model identifier changes. |
Exact-pair decision and evidence guide. Verified 2026-09-01. These three sections are computed from dated scenarios; assumptions and unavailable joins are not observations.
Exact pair boundary: requested models are claude-fable-5 and claude-sonnet-5. A broad provider, pricing, task, or rate-limit verdict is out of scope.
Fable/Sonnet task-consequence admission matrix
Frozen fixture board. Formula / decision rule: eligible = consequence floor + required review + pair coverage; beyond evidence => No supported verdict Boundary: Generic task leadership remains with /best-llm-for pages; this is only the exact Fable/Sonnet admission gate.
| Frozen fixture | Joined fields and evidence | Output | State |
|---|---|---|---|
| routine drafting | consequence=low; review=spot-check; required=factuality/style; Fable/Sonnet coverage=Unresolved | eligible tier=Unavailable until both cells join | UNAVAILABLE — paired coverage gap |
| supervised code | consequence=medium; review=human+tests; required=patch correctness; effort/tools=matched required | Sonnet eligible only after matched evidence clears floor; current result=Unavailable | CONDITIONAL — controlled evidence required |
| autonomous tool loop | consequence=high; review=post-run; required=tool/schema/side-effect controls; coverage=asymmetric | No supported verdict when either model lacks same-loop evidence | NO SUPPORTED VERDICT — asymmetric coverage |
| security review | consequence=high; review=expert; required=recall+false-positive ceiling; pair evidence=missing | do not select Fable from a provider-only claim | NO SUPPORTED VERDICT — expert denominator missing |
| legal-summary support | consequence=high; review=professional; required=source fidelity; model result=Unavailable | route remains human-reviewed; tier=Unavailable | UNAVAILABLE — no admissible pair result |
| irreversible-action planning | consequence=critical; review=operator gate; required=side-effect containment+rollback | neither model receives an autonomous admission | BLOCKED — human/operator gate required |
Provenance: claude-fable-5-vs-claude-sonnet-5, first-party evidence, surface verification date 2026-09-01. Anthropic Claude model documentation. Missing joins fail closed.
Matched-setting evaluation comparability ledger
Frozen fixture board. Formula / decision rule: comparable = same benchmark + version + effort + tools + judge + denominator; incomparable rows cannot be averaged Boundary: A provider-only Fable result or a changed tool policy cannot be used as a Sonnet delta.
| Frozen fixture | Joined fields and evidence | Output | State |
|---|---|---|---|
| same benchmark/version/effort | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=benchmark=v; effort=same; tools=same; judge=same; n=same; evidence=2026-09-01 | delta=Unavailable until both values and denominator are present | CONDITIONAL — comparable shape, values missing |
| different effort | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=benchmark=v; effort=Fable high/Sonnet medium; tools=same; evidence=2026-09-01 | not comparable; preserve rows without winner | NOT COMPARABLE — effort confounder |
| different tool access | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=benchmark=v; effort=same; tools=Fable+search/Sonnet none; evidence=2026-09-01 | not comparable; tool advantage is not model-only | NOT COMPARABLE — tool confounder |
| different benchmark revision | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=benchmark=revA/revB; effort=same; judge=same; evidence=2026-09-01 | delta=Unavailable; revision change blocks averaging | NOT COMPARABLE — revision mismatch |
| provider-only Fable row | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=Fable=provider claim; Sonnet=missing; denominator=missing; evidence=2026-09-01 | Fable value is retained as provider claim, not bilateral evidence | UNAVAILABLE — provider-only row |
| missing Sonnet cell | provider=Anthropic; host=api.anthropic.com; old=claude-fable-5; new=claude-sonnet-5; setting=benchmark=v; Fable=value; Sonnet=missing; sample=Fable only; evidence=2026-09-01 | do not infer Sonnet or compute a delta | FAIL CLOSED — missing side |
Provenance: claude-fable-5-vs-claude-sonnet-5, first-party evidence, surface verification date 2026-09-01. Anthropic Claude model documentation. Missing joins fail closed.
Marginal-value justification board
Frozen fixture board. Formula / decision rule: justified = incremental value × avoided-failure delta > incremental model/reviewer cost; missing success delta => Unavailable Boundary: Rates are linked to pricing owners; task value, retry, and reviewer deltas are assumptions unless locally observed.
| Frozen fixture | Joined fields and evidence | Output | State |
|---|---|---|---|
| low-value batch | value_per_item=assumption; tokens=declared; Fable/Sonnet rates=dated links; success_delta=missing | justification=Unavailable | UNAVAILABLE — success delta missing |
| support workflow | value=declared; volume=declared; output tokens=declared; retry_delta=assumption; review_delta=assumption | incremental cost and avoided failures remain scenario-only | ASSUMPTION ONLY — validate on support set |
| supervised coding | value=reviewer hour; tokens=declared; test_pass_delta=local measurement required; rates=dated | Fable justified only if avoided reviewer/rework cost clears premium | RULE — threshold, no observed verdict |
| agent retry | value=side-effect avoidance; attempt_cap=declared; retry_delta=missing; tools=same required | do not claim Fable value without pair-specific retry evidence | UNAVAILABLE — retry delta missing |
| expensive reviewer hour | value=reviewer hourly cost=assumption; review minutes=assumption; success delta=missing | required avoided failures=formula output; numeric answer=Unavailable | UNAVAILABLE — review inputs missing |
| missing success delta | value=declared; success_delta=missing; Fable premium=rate-owner join; Sonnet baseline=not joined | no justified/not-justified verdict | UNAVAILABLE — no success comparison |
Provenance: claude-fable-5-vs-claude-sonnet-5, first-party evidence, surface verification date 2026-09-01. Anthropic pricing. Missing joins fail closed.
Method and limitations: calculations use only the displayed deterministic rule and visibly labeled assumptions. Missing provider, host, alias, version, effort, tool, modality, benchmark, workload, rate-period, or date joins fail closed. Run this scenario →
Exact-pair decision guide · verified 2026-09-07
Exact pair boundary: Anthropic; same provider, premier creative reasoning vs agentic workhorse, and 5x pricing ratio must join. Requested models are claude-fable-5 and claude-sonnet-5. Provider, rate, pricing, and task pages remain fact owners.
Token pricing and 5x tariff multiple expenditure matrix
Frozen scenario board. Formula / deterministic rule: cost_ratio = fable_cost / sonnet_cost; base tariff 5.0x multiple applies Boundary: Owns base token expenditure comparisons and budget allocation modeling.
| Frozen scenario | Pair, identity, host, surface, and evidence fields | Result | State |
|---|---|---|---|
| 10K marketing copy generation calls | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=10K marketing copy generation calls; workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 10K marketing copy generation calls is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| 50K customer narrative personalization passes | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=50K customer narrative personalization passes; workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 50K customer narrative personalization passes is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| 100K automated document drafts | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=100K automated document drafts; workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 100K automated document drafts is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| prompt caching active (90% read discount both) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=prompt caching active (90% read discount both); workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — prompt caching active (90% read discount both) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| batch processing active (50% both) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=batch processing active (50% both); workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — batch processing active (50% both) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| unresolved pricing currency | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=unresolved pricing currency; workload; prompt tokens; completion tokens; Fable 5 spend; Sonnet 5 spend; budget multiple; pricing champion; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — unresolved pricing currency has no matched, dated bilateral observation. | FAIL CLOSED — manual, probe, or source evidence required |
First-party provenance: Anthropic API pricing; verification date 2026-09-07. Missing or conflicting joins fail closed.
Context window scaling (1M vs 200K) and ingestion gate
Frozen scenario board. Formula / deterministic rule: eligible = (context <= 200K ? Both : (context <= 1M ? Fable : Excluded)) Boundary: Owns context scaling and capability gating between 200K and 1M tokens.
| Frozen scenario | Pair, identity, host, surface, and evidence fields | Result | State |
|---|---|---|---|
| short marketing brief (5K tokens) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=short marketing brief (5K tokens); context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — short marketing brief (5K tokens) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| comprehensive product manual (50K tokens) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=comprehensive product manual (50K tokens); context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — comprehensive product manual (50K tokens) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| full enterprise knowledge base (200K boundary) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=full enterprise knowledge base (200K boundary); context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — full enterprise knowledge base (200K boundary) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| comprehensive legal discovery packet (600K tokens) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=comprehensive legal discovery packet (600K tokens); context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — comprehensive legal discovery packet (600K tokens) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| 1M full context saturation test | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=1M full context saturation test; context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 1M full context saturation test is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| unsupported archive payload format | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=unsupported archive payload format; context size; Fable 5 eligible; Sonnet 5 eligible; Fable cost; Sonnet cost; context gating verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — unsupported archive payload format has no matched, dated bilateral observation. | FAIL CLOSED — manual, probe, or source evidence required |
First-party provenance: Anthropic Claude model documentation; verification date 2026-09-07. Missing or conflicting joins fail closed.
Agent tool calling reliability vs creative prose trade-off
Frozen scenario board. Formula / deterministic rule: break_even_value = fable_spend - sonnet_spend; quality gain must offset cost delta Boundary: Owns accuracy-versus-creative-flair trade-off evaluation.
| Frozen scenario | Pair, identity, host, surface, and evidence fields | Result | State |
|---|---|---|---|
| autonomous coding agent tool loop (Sonnet optimal) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=autonomous coding agent tool loop (Sonnet optimal); workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — autonomous coding agent tool loop (Sonnet optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| creative fiction prose synthesis (Fable optimal) | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=creative fiction prose synthesis (Fable optimal); workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — creative fiction prose synthesis (Fable optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| hybrid technical documentation drafting | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=hybrid technical documentation drafting; workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — hybrid technical documentation drafting is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| customer support conversational agent | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=customer support conversational agent; workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — customer support conversational agent is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| agent loop runaway circuit breaker | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=agent loop runaway circuit breaker; workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — agent loop runaway circuit breaker is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
| unresolved quality evaluation metric | pair=claude-fable-5 vs claude-sonnet-5; provider=Anthropic; hostA=api.anthropic.com; hostB=api.anthropic.com; surfaceA=Anthropic Messages API; surfaceB=Anthropic Messages API; requested/effective IDs=claude-fable-5,claude-sonnet-5; scenario=unresolved quality evaluation metric; workload domain; Sonnet 5 fit; Fable 5 fit; cost comparison; primary selection factor; architectural verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — unresolved quality evaluation metric has no matched, dated bilateral observation. | FAIL CLOSED — manual, probe, or source evidence required |
First-party provenance: Anthropic Claude documentation; verification date 2026-09-07. Missing or conflicting joins fail closed.
Method and limitations: formulas are deterministic; observed and assumed inputs are labeled; no missing provider, host, account, region, realm, alias, snapshot, revision, weight, artifact, control, tool, modality, workload, rate-period, timestamp, or result is transferred. Run this scenario →
How do Claude Fable 5 and Claude Sonnet 5 compare on specs?
| Claude Fable 5 | Claude Sonnet 5 | |
|---|---|---|
| Price (input) | $10.00/M | $2.00/M ✓ |
| Price (output) | $50.00/M | $10.00/M ✓ |
| Blended price | $20.00/M | $4.00/M ✓ |
| Context window | 1,000,000 tokens ✓ | 500,000 tokens |
| Max output | 128,000 tokens ✓ | 64,000 tokens |
| Modalities | text, vision | text, vision |
| Reasoning mode | Yes | Yes |
| Released | 2026-05 | 2026-06 |
| Speed | 41 t/s | Not measured |
How much do Claude Fable 5 and Claude Sonnet 5 cost at scale?
| Tokens / month | Claude Fable 5 | Claude Sonnet 5 | Delta |
|---|---|---|---|
| 1,000,000 | $20.00 | $4.00 | $16.00 (5.0×) |
| 10,000,000 | $200.00 | $40.00 | $160.00 (5.0×) |
| 100,000,000 | $2000.00 | $400.00 | $1600.00 (5.0×) |
Choose Claude Fable 5 if…
- ✓Most capable widely-released Claude of the 5.0 era
- ✓1M-token context
- ✓Extended-thinking mode for hard reasoning
- ✓Legacy ultra-tier Claude workloads superseded by Fable 5.1.
Choose Claude Sonnet 5 if…
- ✓Frontier intelligence at Sonnet economics
- ✓Fast interactive coding performance
- ✓Reliable structured output and extended thinking
- ✓Production workloads that need near-Opus quality with better speed and cost.
Run this exact matchup right now
Send the same prompt to Claude Fable 5 and Claude Sonnet 5 side by side and see the outputs yourself.
Try Claude Fable 5 vs Claude Sonnet 5 FreeWhat are common questions about Claude Fable 5 and Claude Sonnet 5?
Is Claude Fable 5 cheaper than Claude Sonnet 5?
Claude Sonnet 5 is cheaper, at $4.00 per million blended tokens vs $20.00 for Claude Fable 5.
Which has the bigger context window, Claude Fable 5 or Claude Sonnet 5?
Claude Fable 5 has the larger context window: 1,000,000 tokens vs 500,000.
Can Claude Fable 5 replace Claude Sonnet 5 for coding?
Both are viable for coding. Most capable widely-released Claude of the 5.0 era (Claude Fable 5) vs Frontier intelligence at Sonnet economics (Claude Sonnet 5) — pick based on which strength matters more for your workload.
Neither of these? See Claude Sonnet 5 alternatives.
What related comparisons help choose between Claude Fable 5 and Claude Sonnet 5?
Claude Fable 5 pricingClaude Sonnet 5 pricingvs Claude Opus 4.8vs GPT-5.6 Solvs DeepSeek V4 Provs Gemini 3.1 ProPremium model tests
Pricing verified 2026-08-14. Specs verified 2026-08-14.