Skip to content
Mock data. Run values, controls, targets and reproductions on this preview are invented for layout. Claim-card text, upstream figures, the reference price and the RFC list come from the repository.
Inference ProofsMenu

Prover

FlashProve

One proof covers one request: a pass over a 1,024-token window, stating the next token (PRD §1).

Pinned at anthropics/FlashProve@9e2bdd8ff5bf9c62f18263d7dac189b0400b260c

Claim card

What a verifier accept meansconfiguration exact-match-leancard v1draft, not reviewed

The output token is the top choice of the integer model whose weights were committed under the given commitment ID, on the given 1,024-token window.

Shape only, not a reviewed card: the values are the RFC-0002 §Example, to be filled from upstream documentation.

Claim class
next-token/exact-match/committed-weights
Security notion
knowledge_soundness, for the Lean verifier; C++ verifier tested against it
Security level
q/2^124 + 2^-101 after q hash queries, exact-match policy
Model
random oracle, adversary classical
Setup
not stated
Zero knowledge
claimed true; hidden: weights; not hidden: a declared list of weights; status proved_and_tested
Coverage
phase prefill, unit request; tokens covered: the window's 1,024 positions; operations: not stated; policy: exact_match: top token of the integer model
Binding
not stated
Assumptions
  • configuration files fixed before the hash function is drawn
  • the commitment ID is obtained from a trusted party
Residual
proves the integer model, which differs from the 16-bit model's top token at some positions
Availability
research
Forbidden flags
none listed
Reviewed by
no review yet

Rows

One block per workload and configuration. Repetition strips use each metric's own axis.

Workload fp-maverick-w1024-b1

flashprove / exact-match-lean on fp-maverick-w1024-b1

profile prover-h100tier reported
Measured

setup_seconds

118.3 s

scope setupgrade measuredtier reportedmock

mean 118.3, sd 0.7024, min 117.6, max 119, n 3, n_failed 0

mean 118.3 srepetition 1: 118.2 srepetition 2: 117.6 srepetition 3: 119 ssource: mock harness record

commit_seconds

2,406 s

scope commitgrade measuredtier reportedmock

mean 2,406, sd 6.431, min 2,400, max 2,413, n 3, n_failed 0

mean 2,406 srepetition 1: 2405.1 srepetition 2: 2412.6 srepetition 3: 2399.8 ssource: mock harness record

prove_session_seconds

301.5 s

scope prove.sessiongrade measuredtier reportedmock

mean 301.5, sd 1.179, min 300.5, max 302.8, n 3, n_failed 0

mean 301.5 srepetition 1: 301.2 srepetition 2: 302.8 srepetition 3: 300.5 ssource: mock harness record

prove_steady_seconds

234.9 s

scope prove.steadygrade measuredtier reportedmock

mean 234.9, sd 1.106, min 233.9, max 236.1, n 3, n_failed 0

mean 234.9 srepetition 1: 234.8 srepetition 2: 236.1 srepetition 3: 233.9 ssource: mock harness record

whole_wall_seconds

2,707 s

scope whole_wallgrade measuredtier reportedmock

mean 2,707, sd 7.603, min 2,700, max 2,715, n 3, n_failed 0

mean 2,707 srepetition 1: 2706.2999999999997 srepetition 2: 2715.4 srepetition 3: 2700.3 ssource: mock harness record

tokens_covered_total

5,120 token

grade measuredtier reportedmock

mean 5,120, sd 0, min 5,120, max 5,120, n 3, n_failed 0

source: mock harness record

tokens_covered_steady

4,096 token

grade measuredtier reportedmock

mean 4,096, sd 0, min 4,096, max 4,096, n 3, n_failed 0

source: mock harness record

requests_count

5 count

grade measuredtier reportedmock

mean 5, sd 0, min 5, max 5, n 3, n_failed 0

source: mock harness record

proofs_count

5 count

grade measuredtier reportedmock

mean 5, sd 0, min 5, max 5, n 3, n_failed 0

source: mock harness record

proof_bytes_total

93,910,000 B

grade measuredtier reportedmock

mean 93,910,000, sd 30,000, min 93,880,000, max 93,940,000, n 3, n_failed 0

mean 93,910,000 Brepetition 1: 93910000 Brepetition 2: 93880000 Brepetition 3: 93940000 Bsource: mock harness record

verify_seconds

3.113 s

scope verifygrade measuredtier reportedmock

mean 3.113, sd 0.02517, min 3.090, max 3.140, n 3, n_failed 0

mean 3.113 srepetition 1: 3.11 srepetition 2: 3.14 srepetition 3: 3.09 ssource: mock harness record

verify_seconds_total

15.60 s

scope verifygrade measuredtier reportedmock

mean 15.60, sd 0.1000, min 15.50, max 15.70, n 3, n_failed 0

mean 15.60 srepetition 1: 15.6 srepetition 2: 15.7 srepetition 3: 15.5 ssource: mock harness record

verify_single_thread_seconds

21.13 s

scope verify.single_threadgrade measuredtier reportedmock

mean 21.13, sd 0.03606, min 21.10, max 21.17, n 3, n_failed 0

mean 21.13 srepetition 1: 21.12 srepetition 2: 21.17 srepetition 3: 21.1 ssource: mock harness record
Derived (computed by the derivation code)

rate_whole_wall

1.891 token/s

scope whole_wallgrade derivedtier reportedmock

mean 1.891, sd 0.005308, min 1.886, max 1.896, n 3, n_failed 0

mean 1.891 token/srepetition 1: 1.8918819051842 token/srepetition 2: 1.8855417249760624 token/srepetition 3: 1.8960856201162832 token/ssource: mock derivation record

rate_prove.session

16.98 token/s

scope prove.sessiongrade derivedtier reportedmock

mean 16.98, sd 0.06631, min 16.91, max 17.04, n 3, n_failed 0

mean 16.98 token/srepetition 1: 16.99867197875166 token/srepetition 2: 16.908850726552178 token/srepetition 3: 17.03826955074875 token/ssource: mock derivation record

rate_prove.steady

17.43 token/s

scope prove.steadygrade derivedtier reportedmock

mean 17.43, sd 0.08201, min 17.35, max 17.51, n 3, n_failed 0

mean 17.43 token/srepetition 1: 17.444633730834752 token/srepetition 2: 17.348581109699282 token/srepetition 3: 17.51175716117999 token/ssource: mock derivation record

cost_per_1k_tokens_whole_wall

0.6301 USD/1k_token

scope whole_wallgrade derivedtier reportedmock

mean 0.6301, sd 0.001770, min 0.6285, max 0.6320, n 3, n_failed 0

mean 0.6301 USD/1k_tokenrepetition 1: 0.62988427734375 USD/1k_tokenrepetition 2: 0.6320022786458334 USD/1k_tokenrepetition 3: 0.6284877929687501 USD/1k_tokensource: mock derivation record

proof_bytes_per_token

18,340 B/token

grade derivedtier reportedmock

mean 18,340, sd 5.859, min 18,340, max 18,350, n 3, n_failed 0

mean 18,340 B/tokenrepetition 1: 18341.796875 B/tokenrepetition 2: 18335.9375 B/tokenrepetition 3: 18347.65625 B/tokensource: mock derivation record

verify_seconds_per_token

0.003047 s/token

scope verifygrade derivedtier reportedmock

mean 0.003047, sd 1.953e-5, min 0.003027, max 0.003066, n 3, n_failed 0

mean 0.003047 s/tokenrepetition 1: 0.003046875 s/tokenrepetition 2: 0.0030664062499999997 s/tokenrepetition 3: 0.00302734375 s/tokensource: mock derivation record
Runs
  • mock-fp-b1-r1completed
  • mock-fp-b1-r2completed
  • mock-fp-b1-r3completed
  • Superseded, failed and ineligible runs never contribute to statistics (RFC-0009 S7).
Run anatomy
setup
commit
prove.session
prove.steady
whole_wall
verify
releases
Repetition 1, mock values, seconds from the setup call start; total 2,840 s. Scopes as RFC-0004 defines them.

Workload fp-maverick-w1024-b24

flashprove / exact-match-lean on fp-maverick-w1024-b24

profile prover-h100tier reproduced
Measured

setup_seconds

120.7 s

scope setupgrade measuredtier reproducedmock

mean 120.7, sd 0.7506, min 119.9, max 121.4, n 3, n_failed 0

mean 120.7 srepetition 1: 121.4 srepetition 2: 119.9 srepetition 3: 120.7 ssource: mock harness record

commit_seconds

2,411 s

scope commitgrade measuredtier reproducedmock

mean 2,411, sd 12.35, min 2,398, max 2,423, n 3, n_failed 0

mean 2,411 srepetition 1: 2410.5 srepetition 2: 2398.2 srepetition 3: 2422.9 ssource: mock harness record

prove_session_seconds

1,186 s

scope prove.sessiongrade measuredtier reproducedmock

mean 1,186, sd 6.304, min 1,179, max 1,192, n 3, n_failed 0

mean 1,186 srepetition 1: 1185.3 srepetition 2: 1192 srepetition 3: 1179.4 ssource: mock harness record

prove_steady_seconds

887.9 s

scope prove.steadygrade measuredtier reproducedmock

mean 887.9, sd 6.678, min 881.6, max 894.9, n 3, n_failed 0

mean 887.9 srepetition 1: 887.2 srepetition 2: 894.9 srepetition 3: 881.6 ssource: mock harness record

whole_wall_seconds

3,596 s

scope whole_wallgrade measuredtier reproducedmock

mean 3,596, sd 6.056, min 3,590, max 3,602, n 3, n_failed 0

mean 3,596 srepetition 1: 3595.8 srepetition 2: 3590.2 srepetition 3: 3602.3 ssource: mock harness record

tokens_covered_total

122,900 token

grade measuredtier reproducedmock

mean 122,900, sd 0, min 122,900, max 122,900, n 3, n_failed 0

source: mock harness record

tokens_covered_steady

98,300 token

grade measuredtier reproducedmock

mean 98,300, sd 0, min 98,300, max 98,300, n 3, n_failed 0

source: mock harness record

requests_count

120 count

grade measuredtier reproducedmock

mean 120, sd 0, min 120, max 120, n 3, n_failed 0

source: mock harness record

proofs_count

120 count

grade measuredtier reproducedmock

mean 120, sd 0, min 120, max 120, n 3, n_failed 0

source: mock harness record

proof_bytes_total

2.254e9 B

grade measuredtier reproducedmock

mean 2.254e9, sd 602,800, min 2.253e9, max 2.254e9, n 3, n_failed 0

mean 2.254e9 Brepetition 1: 2253600000 Brepetition 2: 2254100000 Brepetition 3: 2252900000 Bsource: mock harness record

verify_seconds

3.120 s

scope verifygrade measuredtier reproducedmock

mean 3.120, sd 0.03000, min 3.090, max 3.150, n 3, n_failed 0

mean 3.120 srepetition 1: 3.12 srepetition 2: 3.09 srepetition 3: 3.15 ssource: mock harness record

verify_seconds_total

374.4 s

scope verifygrade measuredtier reproducedmock

mean 374.4, sd 3.600, min 370.8, max 378, n 3, n_failed 0

mean 374.4 srepetition 1: 374.4 srepetition 2: 370.8 srepetition 3: 378 ssource: mock harness record

verify_single_thread_seconds

21.14 s

scope verify.single_threadgrade measuredtier reproducedmock

mean 21.14, sd 0.03000, min 21.11, max 21.17, n 3, n_failed 0

mean 21.14 srepetition 1: 21.14 srepetition 2: 21.11 srepetition 3: 21.17 ssource: mock harness record
Derived (computed by the derivation code)

rate_whole_wall

34.17 token/s

scope whole_wallgrade derivedtier reproducedmock

mean 34.17, sd 0.05753, min 34.11, max 34.23, n 3, n_failed 0

mean 34.17 token/srepetition 1: 34.173202069080595 token/srepetition 2: 34.22650548715949 token/srepetition 3: 34.11153984954057 token/ssource: mock derivation record

rate_prove.session

103.6 token/s

scope prove.sessiongrade derivedtier reproducedmock

mean 103.6, sd 0.5510, min 103.1, max 104.2, n 3, n_failed 0

mean 103.6 token/srepetition 1: 103.66995697291826 token/srepetition 2: 103.08724832214764 token/srepetition 3: 104.18857045955569 token/ssource: mock derivation record

rate_prove.steady

110.7 token/s

scope prove.steadygrade derivedtier reproducedmock

mean 110.7, sd 0.8317, min 109.8, max 111.5, n 3, n_failed 0

mean 110.7 token/srepetition 1: 110.80252479711451 token/srepetition 2: 109.84914515588333 token/srepetition 3: 111.50635208711434 token/ssource: mock derivation record

cost_per_1k_tokens_whole_wall

0.03487 USD/1k_token

scope whole_wallgrade derivedtier reproducedmock

mean 0.03487, sd 5.873e-5, min 0.03482, max 0.03493, n 3, n_failed 0

mean 0.03487 USD/1k_tokenrepetition 1: 0.03487137858072917 USD/1k_tokenrepetition 2: 0.03481707085503472 USD/1k_tokenrepetition 3: 0.034934414333767365 USD/1k_tokensource: mock derivation record

proof_bytes_per_token

18,340 B/token

grade derivedtier reproducedmock

mean 18,340, sd 4.905, min 18,330, max 18,340, n 3, n_failed 0

mean 18,340 B/tokenrepetition 1: 18339.84375 B/tokenrepetition 2: 18343.912760416668 B/tokenrepetition 3: 18334.147135416668 B/tokensource: mock derivation record

verify_seconds_per_token

0.003047 s/token

scope verifygrade derivedtier reproducedmock

mean 0.003047, sd 2.930e-5, min 0.003018, max 0.003076, n 3, n_failed 0

mean 0.003047 s/tokenrepetition 1: 0.0030468749999999997 s/tokenrepetition 2: 0.0030175781250000003 s/tokenrepetition 3: 0.003076171875 s/tokensource: mock derivation record
Runs
  • mock-fp-b24-r1completed
  • mock-fp-b24-r2completed
  • mock-fp-b24-r3completed
  • Superseded, failed and ineligible runs never contribute to statistics (RFC-0009 S7).
Run anatomy
setup
commit
prove.session
prove.steady
whole_wall
verify
releases
Repetition 1, mock values, seconds from the setup call start; total 4,092 s. 120 proofs released; the first 60 release ticks are drawn. Scopes as RFC-0004 defines them.

Controls

All controls

fp-maverick-w1024-b1 profile prover-h100

6 passed0 failed4 void, not applicable or error

fp-maverick-w1024-b24 profile prover-h100

6 passed0 failed4 void, not applicable or error

Reference rows on the same workloads

RFC-0015

ref-serve on fp-maverick-w1024-b24

profile prover-h100tier reported
Measured

whole_wall_seconds

null (not_measurable_on_profile) s

scope whole_wallgrade measuredtier reportedmock
source: mock harness record
Derived (computed by the derivation code)

rate_whole_wall

null (not_measurable_on_profile) token/s

scope whole_wallgrade derivedtier reportedmock
source: mock derivation record
Runs

No runs: the reference system cannot run this workload on the profile.

Figures the system publishes upstream

Real data from systems/flashprove/external.yaml, grade external

Copied at the pin, never recomputed or rounded in the record. They are context, never inputs to a measured number (RFC-0004 INV-MAP-1). The mapping from the upstream scope is marked approximate where the definitions differ.

Upstream figures
MetricWorkloadValue and labelsUpstream scopeNote
rate_prove.steadyfp-maverick-w1024-b24

111.2 token/s

scope prove.steadygrade externaltier reported
source: upstream
Proving speed in steady service, 24 requests a batch mapping approximatemean of 3 runs; sd 0.92; 110.20 to 111.99; public 4-bit files
rate_prove.sessionfp-maverick-w1024-b24

103.6 token/s

scope prove.sessiongrade externaltier reported
source: upstream
Speed counted from the program's start to the last release, five batches only mapping approximatemean of 3 runs; sd 0.84; 102.65 to 104.24; public 4-bit files
rate_prove.steadyfp-maverick-w1024-b1

17.41 token/s

scope prove.steadygrade externaltier reported
source: upstream
Proving speed in steady service, 1 request a batch mapping approximatemean of 3 runs; sd 0.15; 17.24 to 17.51; public 4-bit files
verify_single_thread_secondsfp-maverick-w1024-b24

21.14 s

scope verify.single_threadgrade externaltier reported
source: upstream
The verifier on one proof with one thread mapping approximatemean of 3 runs; sd 0.03; 21.11 to 21.17

Upstream pin anthropics/FlashProve@9e2bdd8ff5bf9c62f18263d7dac189b0400b260c

Reply from the system maintainers

No reply for this release. System maintainers may write up to 300 words, shown here unedited (RFC-0009 §Right of reply).