🎁 New Sign up free, 10 calls on us. Up to $1, no card needed.

Claude Fable 5 vs Kimi K3

vs

Which one, when — curated verdict, not a benchmark table

kimi-k3 undercuts claude-fable-5 by about 3.3x across the rate card ($3 vs $10 per million input, $15 vs $50 output, $0.3 vs $1 cached reads), and it also accepts video input and can emit up to 1,048,576 output tokens inside its 1,048,576-token window. claude-fable-5 offers a 1,000,000-token context but caps output at 128,000 tokens, and it carries an explicit thinking capability plus a January 2026 knowledge cutoff. Pick kimi-k3 for video input and very long generations at lower cost; pick claude-fable-5 when that thinking capability and stated recency matter more than price.

Pricing

Claude Fable 5 Kimi K3 Δ
Input / 1M tokens $10 $3 3.3×
Output / 1M tokens $50 $15 3.3×
Cache read / 1M tokens $1 $0.3 3.3×
Cache write 1.25x (5m) / 2x (1h)

Rates from the live catalog at build time; each model page carries the current card.

Where they sit — input price per 1M tokens across all 63 chat models on this billing unit (log scale)

Claude Fable 5 · $10 Kimi K3 · $3
$0.05 · Qwen3 VL Flash $30 · GPT-5.4 Pro

Capabilities

Claude Fable 5 Kimi K3
Tool use yes yes
Thinking control always on always on
Structured output yes yes
Prompt caching explicit (you mark the prefix) implicit (automatic)
Cache lifetime 5m default, 1h option not published
Minimum cached prefix 1024 tokens not published

Specs

Claude Fable 5 Kimi K3
Input modalities text image text image video
Output modalities text text
Released 2026-06-09
Knowledge cutoff 2026-01
Context window 1M 1M
Max output 128K 1M
Thinking parameter output_config.effort (thinking.type is adaptive-only and needs no configuration) reasoning_effort (top-level; the thinking object is not accepted)
Accepted values
effort
  • low
  • medium
  • high
  • xhigh
  • max

both "enabled" and "disabled" return 400

reasoning_effort
  • low
  • high
  • max
Default

thinking always on (adaptive)

effort
  • high
max

Specs are transcribed from each vendor’s documentation; a row a vendor does not publish is left out rather than inferred. Full sources: Claude Fable 5 · Kimi K3

One prompt, both models — measured through the gateway

PROMPT Explain to a CFO, in exactly three sentences, why our LLM bill doubled after we enabled reasoning. CHECK exactly 3 sentences

Claude Fable 5 passed · 3 sentences

When we enabled reasoning, the model began generating extensive internal "thinking" tokens before each answer, and those tokens are billed as output even though users never see them. Output tokens typically cost several times more than input tokens, so adding thousands of hidden reasoning tokens per request effectively multiplied our per-query cost. In short, we're paying for a much longer response behind the scenes in exchange for higher answer quality, and we can manage this by tuning reasoning effort levels or reserving reasoning mode for complex tasks only.

out 228 tok latency 6.7 s

Kimi K3 passed · 3 sentences

Reasoning models don't just answer questions—they "think" first, generating long internal chains of step-by-step logic before producing a response. Those hidden thinking steps are billed as output tokens (the most expensive kind, typically 3–5x the price of input tokens), and a single query can generate thousands of them even when the visible answer is only a paragraph long. So you're paying for dramatically more compute per request: the bill doubled because the model does far more work behind the scenes, not because usage increased.

out 755 tok (+637 thinking) latency 20.8 s

Instruction following (exactly three sentences — countable), audience shaping (CFO register), and the hidden-thinking billing gap the token meters below expose.

PROMPT This function is supposed to return the longest run of consecutive equal items, but callers report it is off by one on single-element inputs and misses runs that end at the last index. Fix it and explain the bug in one paragraph. CHECK fix passes tests

Claude Fable 5 passed · 8/8 cases

## Fixed code ```python def longest_run(xs): if not xs: return 0 best = cur = 1 for i in range(1, len(xs)): if xs[i] == xs[i - 1]: cur += 1 else: cur = 1 best = max(best, cur) return best ``` ## Explanation The original function counts **adjacent equal pairs** rather than **items in a run**: a run of length *n* contains only *n − 1* equal neighbor pairs, so every result is off by one — a single-element list yields 0 instead of 1, and `[1, 2, 2]` yields 1 instead of 2 (which callers perceive as "missing" the run at the end). The fix is to initialize both `cur` and `best` to 1 (any nonempty list has a run of at least length 1), reset `cur` to 1 rather than 0 when

out 531 tok latency 12.5 s

Kimi K3 passed · 8/8 cases

```python def longest_run(xs): if not xs: return 0 best = 1 cur = 1 for i in range(1, len(xs)): if xs[i] == xs[i - 1]: cur += 1 best = max(best, cur) else: cur = 1 return best ``` The bug is a fencepost error: `cur` counts *matching adjacent pairs* rather than *items in the run*, because it starts at 0 and only increments when `xs[i] == xs[i - 1]`. A run of length L contains L−1 equal pairs, so every run was undercounted by exactly one. For a single-element list (a run of length 1) the loop never runs and the function returns 0 instead of 1; likewise any run ending at the last index — e.g. the two `2`s in `[1, 2, 2]` — reported 1 instead of 2, making it look like no run existed at all. […]

out 1837 tok (+1547 thinking) latency 47.2 s

Whether the fix is actually correct (runnable), explanation density, and token efficiency on a bounded task.

PROMPT Extract a JSON object with fields {company, quarter, revenue_usd_m, growth_pct, guidance} from this text. Use null for anything not stated; add no other fields. CHECK valid JSON, schema exact

Claude Fable 5 passed · 5/5 fields, guidance null

```json { "company": "Meridian Robotics", "quarter": "Q3 (September quarter)", "revenue_usd_m": 412, "growth_pct": 11, "guidance": null } ``` Notes on interpretation: "up eleven and change" is rendered as 11 (the precise figure isn't stated), and guidance is null because it was explicitly withheld.

out 192 tok latency 6.5 s

Kimi K3 passed · 5/5 fields, guidance null

{ "company": "Meridian Robotics", "quarter": "September quarter", "revenue_usd_m": 412, "growth_pct": 11, "guidance": null }

out 924 tok (+863 thinking) latency 25.9 s

Schema obedience (no invented fields), hallucination pressure (guidance is explicitly withheld), and structured-output path differences.

PROMPT Write a 120-word product announcement for an API gateway feature that caches prompts across providers. Forbidden words: "seamless", "unlock", "game-changer", "revolutionize", "empower". Exactly one sentence must be a question. CHECK 120 words, 0 banned words

Claude Fable 5 passed · 120 words, 0 banned, 1 question

**Introducing Cross-Provider Prompt Caching for our API Gateway** Starting today, your repeated prompts are cached once and reused across OpenAI, Anthropic, Google, and any other provider you route through the gateway. Instead of paying full token costs every time your application sends the same system prompt or context block, the gateway detects matching prefixes and serves cached results in milliseconds. Why keep spending money and latency on identical requests? Early adopters report cost reductions of up to 60 percent on high-volume workloads, along with faster median response times. Configuration is simple: enable caching in your dashboard, set a TTL, and choose which routes participate. Cache entries are encrypted at rest and never shared between accounts. Available now on all plans.

out 1173 tok latency 18.2 s

Kimi K3 passed · 120 words, 0 banned, 1 question

Meet Prompt Cache, a new API gateway capability that stores prompt responses and serves them across OpenAI, Anthropic, Google, and Azure endpoints. It matches requests by model, prompt hash, tools, temperature, and tenant policy, so repeated work returns fast while sensitive contexts stay isolated. Teams set TTLs, stale rules, encryption scopes, and bypass flags per route. Analytics show hit rate, latency saved, token spend avoided, and drift risk by provider. What changes for developers? Keep one integration, add cache headers, and watch fallback logic respect consent, residency, and audit needs. During rollout, canary keys compare fresh answers with cached copies before promotion. Prompt Cache cuts vendor calls, steadies p95 latency, and gives platform owners controls for cost, quality, and compliance.

out 1527 tok (+1354 thinking) latency 37.9 s

Constraint obedience (word budget, banned-word list, the single question), style fingerprint, and length control.

Switch between them with one line

Both ids are in every tab below — the highlighted pair of lines is the only edit. Same endpoint, same key, same request shape.

from openai import OpenAI

client = OpenAI(
    base_url="https://synthorai.io/v1",
    api_key="sk-syn-...",
)

resp = client.chat.completions.create(
    model="claude-fable-5",
    # model="kimi-k3",  # uncomment this line, comment the one above
    messages=[{"role": "user", "content": "Summarize this diff"}],
    reasoning_effort="medium",
)
print(resp.choices[0].message.content)

Get an API key →

FAQ

Which is cheaper, Claude Fable 5 or Kimi K3?

Kimi K3 is cheaper on input / 1m tokens ($3 vs $10, 3.3× apart). Other rows may point the other way — the table above carries the full card, and real cost depends on your mix.

Can I A/B test Claude Fable 5 against Kimi K3 without two integrations?

Yes. Both are served through the same OpenAI-compatible endpoint with one API key — switching is a one-line model-string change, so you can route a fraction of traffic to each and compare bills directly.

Do Claude Fable 5 and Kimi K3 support prompt caching?

Yes — both bill cached reads below their input rate, so warm-prefix workloads cost less than the list rates suggest. The exact cache-read rows are in the pricing table above.

Related comparisons

From our measured studies